Tattoo Facts

What the monograph says, and what it never mentions

The classification comes from IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 93: Carbon Black, Titanium Dioxide, and Talc, published by IARC in Lyon in 2010 from a Working Group that met in February 2006. The full 466-page PDF was downloaded and searched. The word “tattoo” appears zero times in it. IARC has never evaluated tattooing, and said so itself in 2024.

The Group 2B label rests on inhalation. Verbatim from section 6.4 of the monograph: “Three studies of female rats that inhaled carbon black and three additional studies of female rats exposed intratracheally found significant increases in the incidence of malignant lung tumours, providing sufficient evidence that carbon black can cause cancer in animals.” The human evidence is lung cancer mortality among carbon black production workers, which the Working Group called inconsistent and inadequate. The formal evaluation reads: “There is inadequate evidence in humans for the carcinogenicity of carbon black.”

What happened when carbon black was actually put into skin

Section 5.3 of the same monograph reports the experiments that matter for a tattoo argument, because they are the ones where carbon black went into or under skin rather than into lungs. Two sentences, verbatim:

Read the second one twice. Same species, same route, same category of material. The carbon black that carried extractable polycyclic aromatic hydrocarbons produced tumours. The carbon black with no detectable polycyclic aromatic hydrocarbon did not. Where carbon black was placed into skin, tumours tracked the contaminant load, not the carbon.

That is the experimental reason the European restriction controls polycyclic aromatic hydrocarbons rather than banning carbon. Under REACH Annex XVII entry 75, benzo[a]pyrene is capped at 0.0000005% by weight, which is five parts per billion and the strictest single limit in the entry. Other polycyclic aromatic hydrocarbons classified as carcinogens or mutagens in categories 1A, 1B or 2 are capped at 0.00005%. The control is aimed at what the animal data implicated.

What survives this section: black tattoo ink is carbon black, carbon black carries a Group 2B classification, and that classification is about breathing carbon black dust in an industrial setting. It has never been an evaluation of tattooing, and the skin experiments inside the same monograph point at contaminant chemistry rather than at carbon.

What Nielsen and colleagues actually found

Nielsen C, Jerkeman M and Jöud AS, “Tattoos as a risk factor for malignant lymphoma: a population-based case-control study,” eClinicalMedicine 2024;72:102649 (PMID 38827888). Swedish register-based case-control design, incident lymphoma 2007 to 2017, ages 20 to 60, three matched controls per case. Total 11,905 invited, with 1,398 cases and 4,193 controls responding, a response rate of 54% and 47%. Tattoo status was collected by questionnaire in 2021, in several cases years after diagnosis. Tattoo prevalence was 21% among cases and 18% among controls.

Note the last two rows. If ink caused lymphoma, more ink should carry more risk. The authors write: “We found no evidence of increasing risk with a larger area of total tattooed body surface.” The smaller-area category carried the higher point estimate. Their own conclusion is unambiguous: “Causality cannot be conferred from a single epidemiologic study and the results need to be confirmed by further research.”

They also list their own limitations, verbatim: tattooed body area was assessed at the time of the 2021 survey rather than at the index year; selection bias could not be formally assessed because individual-level data on non-participants was unavailable; and residual confounding or measurement error in self-reported confounders such as smoking could not be ruled out.

The rest of the primary literature

Both meta-analyses are null

McConnell T and colleagues, eClinicalMedicine 2025;89:103563 (PMID 41140458), pooled 4 observational studies covering 17,941 participants, 2,485 cases and 15,456 controls. Verbatim: “No significant association was found between tattooing and lymphoma.” Odds ratios were 1.01 (0.82 to 1.24) for non-Hodgkin lymphoma, 1.01 (0.77 to 1.33) for follicular lymphoma and 0.89 (0.54 to 1.46) for diffuse large B-cell lymphoma. GRADE certainty: low.

Tudella GCN and colleagues, Clinical and Translational Oncology, online 6 June 2026 (PMID 42250187), pooled 7 observational studies covering 140,841 participants. Skin cancer OR 0.92 (0.83 to 1.04). Haematological malignancies OR 1.02 (0.78 to 1.32). One result cuts the other way and belongs on the page: a sensitivity analysis excluding one influential study produced OR 1.20 (1.05 to 1.39) for haematological malignancies, which is significant. The authors' own wording is that an association “emerged only after sensitivity analysis, warranting cautious interpretation and further prospective investigation.”

What the authoritative bodies say

Which pigments, and how many

The European Commission Joint Research Centre inventory (Safety of tattoos and permanent make-up: Final report, EUR 27947 EN, 2016, full PDF) lists 113 colorants for tattoo inks and 100 for permanent make-up. Verbatim: “the large majority consists of organic colorants (92 and 84 corresponding to 81% and 84% for tattoo and PMU respectively). Among the organic colorants, the chemical class of azo pigments is the most represented.”

The best analytical source for a United States studio is Moseman K and colleagues, Analytical Chemistry 2024;96(9):3906 to 3913 (PMID 38387033), which examined 54 inks from 9 brands by Raman spectroscopy, X-ray fluorescence, nuclear magnetic resonance and gas chromatography with mass spectrometry. Verbatim: “Only ten unique pigments were identified in this study: PR254, PR170, PR210, PR112, PY74, PY14, PG7, PB15, PW6, and Carbon Black.” Titanium dioxide and carbon black were used by all nine manufacturers.

The line from that paper worth carrying on its own: “Out of 54 inks, 45 contained unlisted additives and/or pigments.” Forty-five of fifty-four bottles held something the label did not declare.

A Danish registry study of label data, Serup J and colleagues in Dermatology 2025;241(3):259 to 271, catalogued 10,833 bottles carrying a Colour Index designation. Raw counts, leading entries: Pigment White 6 (titanium dioxide) in 5,814 bottles; Pigment Black 6 (carbon black) in 4,074; Pigment Blue 15 in 2,546; Pigment Yellow 14 in 1,542; Pigment Red 101 in 1,054; Pigment Red 170 in 895; Pigment Green 36 in 844. Only raw counts are quoted here, because that paper's percentage table implies a denominator inconsistent with its own stated total.

Converging summary across all three: black is carbon black; white is titanium dioxide; blues and greens are copper phthalocyanines; warm colours are dominated by azo pigments, confirmed chemically by Schreiver I and colleagues at the BfR in Archives of Toxicology 2016;90(7):1639 to 1650, who characterised 36 purified pigments by pyrolysis gas chromatography with mass spectrometry.

Metals, with the exceedances counted

The current best measurement is Toxics 2025;13(11):934, published 30 October 2025, covering 41 products on the European market (27 tattoo inks, 9 dual-use, 5 permanent make-up) by inductively coupled plasma optical emission spectrometry and related methods.

These were market-surveillance samples, chosen for testing rather than drawn at random from what is sold. Nickel exceeding in 24 of 41 products is a real finding about those 41 products. It is not a statement that most tattoo ink exceeds the limit.

Two corroborating studies, read at abstract level only: Violi and colleagues, J Hazard Mater 2026;502:140874, tested 15 Australian inks and reported that “Eight of thirteen regulated elemental substances … exceeded European Union thresholds in at least one sample” and that “all inks tested would be prohibited under EU regulations.” Wang X and colleagues, Contact Dermatitis 2021;85(3):340 to 353, tested 73 Swedish and Austrian inks and found “Total chromium (0.35-139 µg/g) and nickel (0.1-41 µg/g) … in almost all samples.” Wang and the 2025 study are frequently presented as a trend over time. They are not comparable, because Wang measured against the older Council of Europe limits and the 2025 study measured against the far stricter REACH entry 75.

Polycyclic aromatic hydrocarbons, and the distinction most coverage loses

The JRC analysed 358 inks for total polycyclic aromatic hydrocarbons and found, verbatim, that “43% of the 358 inks analysed for total PAHs content presented concentrations well above the threshold of 0.5 mg/kg suggested by the CoE ResAP(2008)1”, with a measured range of 0.5 to 55,000 mg/kg. For benzo[a]pyrene specifically, across a different sample of 300, “24% of the 300 samples analysed contained this substance in levels higher than the maximum recommended quantity of 0.005 mg/kg.” A Basel-Stadt cantonal laboratory survey of 229 inks in 2014 found 14 samples, 6%, with clearly excessive polycyclic aromatic hydrocarbon levels between 4.1 and 64 mg/kg.

The counterpoint sits in the same JRC report and almost never travels with the statistic. Verbatim: “The fact that 57% of them were compliant proves that it is technically feasible to produce carbon black with low levels of PAHs.” This is a manufacturing-quality problem with a demonstrated solution, not a property of black pigment.

Five percentages, five different denominators

The JRC pooled national market-surveillance campaigns run in different years, and its own headline figures do not share a denominator. Listing them as a series of statements about “tattoo inks” invents a common sample that does not exist.

Two further JRC findings worth stating plainly. Of the 67 azo colorants in use, 31 (46%) “contain and might release, by simple reductive cleavage of the azo bond, one of the amines included in the negative lists”, rising to 44 of 67 if amide hydrolysis is included. And the risk ranking the JRC arrived at, verbatim: “The main risks identified, in descending order, are the presence of PAHs, PAAs, microorganisms, heavy metals and preservatives.”

“Vegan” is a supply-chain claim, not a safety claim

No regulator anywhere defines, certifies or polices the word “vegan” for tattoo ink. The Vegan Society Trademark standard requires no animal product, by-product or derivative, no animal testing, and minimised cross-contamination. It imposes no purity, heavy-metal, polycyclic aromatic hydrocarbon or contaminant requirement, and tattoo ink is not a listed category. Certified Vegan lists five eligible product categories, none of which is tattoo ink.

The animal-derived ingredients are real: the JRC lists Pigment Black 9 (bone charcoal) as a black colorant, and shellac, gelatine and glycerol as auxiliaries. What is undocumented is how common the animal-derived versions are. Glycerol was listed in 36 of the 54 inks Moseman analysed, and no analytical method distinguishes animal-derived from plant-derived glycerol, which is precisely why the claim cannot be verified by testing. A literature search for any study comparing vegan and non-vegan ink safety returned nothing. A vegan ink and a non-vegan ink can carry identical pigments and identical metal impurities.

Schreiver I and colleagues, Scientific Reports 2017;7(1):11395 (PMID 28900193), used synchrotron X-ray fluorescence mapping at the European Synchrotron Radiation Facility in Grenoble, plus infrared microscopy and mass spectrometry, on post-mortem tissue. The sample is 4 tattooed cadaver donors and 2 non-tattooed controls, six people in total, yielding 20 skin and 25 lymph node samples from the tattooed donors. The authors state the constraint themselves: “The sample size was limited by the availability of specimens and the beamtime at ESRF.”

What they demonstrated, verbatim: “simultaneous transport of organic pigments, heavy metals and titanium dioxide from skin to regional lymph nodes. Among these compounds, organic pigments displayed the broadest size range with smallest species preferentially reaching the lymph nodes.”

What they did not claim matters as much. There is no cancer claim anywhere in the paper. There is no claim of clinical harm in these donors; the paper states that “most tattooed individuals including the donors analyzed here do not suffer from chronic inflammation.” There is no claim about distant organs, which the authors flag as future work. Their causal language is hedged to “likely contribute.”

A companion paper from the same laboratory, Particle and Fibre Toxicology 2019;16(1):33 (PMID 31451117), reports a second metal source that has nothing to do with ink: needle wear. Tattoo needles “contain nickel (6-8%) and chromium (15-20%)”, and wear particles were found in skin and lymph nodes. Wear rose significantly when tattooing with titanium dioxide white compared with carbon black. The authors' own hedge is that the impact on allergy formation is “as yet to be assessed”, and the human component is one sensitised patient's biopsy plus a pig-skin experiment.

Both papers come from one laboratory, so they do not independently corroborate each other. Independent clinical evidence that the phenomenon is real and occasionally symptomatic comes from the Copenhagen tattoo clinic series, which lists lymphopathy among its diagnostic categories.

European Union: REACH Annex XVII entry 75

Commission Regulation (EU) 2020/2081 of 14 December 2020 amended Annex XVII to REACH. The EU-adopted English text is the source used here, because every EUR-Lex deep link redirected to the Official Journal homepage and ECHA returned a firewall block.

Entry 75 does not list banned inks. It restricts substances by hazard classification in any mixture used for tattooing, and the definition of “for tattooing purposes” is broad, covering “injection or introduction of the mixture into a person's skin, mucous membrane or eyeball, by any process or procedure (including procedures commonly referred to as permanent make-up, cosmetic tattooing, micro-blading and micro-pigmentation).”

Dates: adopted 14 December 2020, in force on the twentieth day after publication, applying from 4 January 2022, with labelling obligations starting the same day. Two pigments got a derogation until 4 January 2023: Pigment Blue 15:3 and Pigment Green 7. The reason, from recital 15, is that during consultation stakeholders identified them as the only two colorants “essential for tattooing on account of the fact that there were no safer and technically adequate alternatives available.”

A detail nearly every summary gets wrong: the scientific committee SEAC recommended a derogation of 36 months. The Commission cut it to 24, stating in recital 40 that “24 months is sufficient to find safer alternatives and to remove mixtures placed on the market … containing these pigments.” The 24-month figure is what became 4 January 2023.

Entry 75 also puts a duty on the artist rather than the manufacturer: “Before using a mixture for tattooing purposes, the person using the mixture shall provide the person undergoing the procedure with the information marked on the package or included in the instructions for use.” And: “Mixtures that do not contain the statement 'Mixture for use in tattoos or permanent make-up' shall not be used for tattooing purposes.”

United States: what the FDA actually does

From the FDA fact sheet “Tattoos & Permanent Makeup”, content current as of 15 October 2024, verbatim: “FDA considers the inks used in intradermal tattoos, including permanent makeup, to be cosmetics. … The pigments used in the inks are color additives, which are subject to premarket approval under the Federal Food, Drug, and Cosmetic Act. However, because of other competing public health priorities and a previous lack of evidence of safety problems specifically associated with these pigments, FDA traditionally has not exercised regulatory authority for color additives on the pigments used in tattoo inks. The actual practice of tattooing is regulated by local jurisdictions.”

And the sharpest line in the document: “Although a number of color additives are approved for use in cosmetics, none is approved for injection into the skin.” The FDA adds that many pigments used in tattoo inks are not approved for skin contact at all, and that some “are industrial grade colors that are suitable for printers' ink or automobile paint.”

Four statements, all true at once, and garbling any one of them produces a myth: tattoo inks are cosmetics under the Act; the pigments are colour additives; colour additives do legally require premarket approval; and no colour additive has been approved for injection into skin, while the FDA has historically exercised enforcement discretion.

The Modernization of Cosmetics Regulation Act of 2022 (Public Law 117-328, division FF, title III, section 3502) added sections 604 to 614 to the Act. Its concrete, checkable effect on tattoo inks is in the small-business exemption at 21 U.S.C. section 364h, which does not apply to “Cosmetic products that are injected” or to products “intended to alter appearance for more than 24 hours.” Tattoo ink hits both exclusions, so no tattoo ink manufacturer, at any revenue level, can claim the small-business exemption from good manufacturing practice, facility registration or product listing. That reading is taken directly from unambiguous statutory text; no secondary source analysing it specifically for tattoo inks was located.

What the 2022 Act did not do: it did not create premarket approval for cosmetics, and it did not change the colour-additive regime. The position that no colour additive is approved for injection into skin is unchanged.

On the enforcement record, the FDA Constituent Update of 24 October 2024 states verbatim: “Between 2003 and 2024, firms conducted 18 voluntary recalls of tattoo inks that were contaminated with a variety of microorganisms.” The most recent was 23 August 2024, when Sierra Stain LLC recalled three water-based pigments distributed to all 50 states, contaminated with organisms including Citrobacter braakii, Pseudomonas fluorescens and Achromobacter xylosoxidans, with no illnesses reported at the time of the notice. Eighteen recalls across 21 years is a count of regulatory actions, not a rate. No published figure exists for units of tattoo ink sold in the United States, so no contamination rate can be computed from it.

California regulates the chair, not the bottle

California Health and Safety Code, division 104, part 15, chapter 7, sections 119300 to 119328, cited as California's Safe Body Art Act, from leginfo.legislature.ca.gov. Its stated purpose, verbatim, is “to provide minimum statewide standards for the regulation of persons engaged in the business or performance of tattooing, body piercing, branding, and the application of permanent cosmetics”, intended “to protect both the practitioner and the client from transmission of infectious diseases.” Note the scope: infectious disease and cross-contamination. California sets no chemical limits on tattoo ink whatsoever. There is no state analogue to REACH entry 75.

One clause deserves separate billing. Section 119303(a)(4) requires the consent form to include, verbatim: “Notice that tattoo inks, dyes, and pigments have not been approved by the federal Food and Drug Administration and that the health consequences of using these products are unknown.” California takes the federal enforcement gap and converts it into an affirmative written disclosure duty at the chair. Every legally tattooed person in California has been told this in writing.

Santa Monica does not run its own environmental health department, so the local enforcement agency for this address is the Los Angeles County Department of Public Health, which registers body art practitioners and permits facilities under the Act.

Tattooing and piercing are not treated alike for minors, and the difference is frequently reported wrong. Tattooing a person under 18 is a misdemeanor under Penal Code section 653 and parental consent cannot authorise it. Piercing a minor is an infraction under Penal Code section 652 that parental presence cures, with ear piercing expressly excluded and nipple or genital piercing of a minor prohibited outright.

The most defensible number on this page

Two fully independent studies, different countries, different decades, different sampling frames, landed 0.1 percentage points apart.

• 10.3%. Brady BG, Gold H, Leger EA, Leger MC, “Self-reported adverse tattoo reactions: a New York City Central Park study,” Contact Dermatitis 2015;73(2):91 to 99 (PMID 26016445). Denominator 300 randomly selected tattooed people. Verbatim: “Of 300 participants, 31 (10.3%) reported experiencing an adverse tattoo reaction, 13 (4.3%) reported acute reactions, and 18 (6.0%) suffered from a chronic reaction involving a specific colour lasting for >4 months.”
• 10.2%. Friis K and colleagues, Dermatology 2024;240(2):297 to 303 (PMID 38081147). Danish population survey of 33,925 respondents aged 16 and over, of whom 5,914 were tattooed. Verbatim: “Among participants with at least one tattoo, 10.2% had experienced tattoo-associated skin reactions beyond the first 3 weeks after their tattoo was made.” Larger tattoos and tattoos held more than 10 years raised the odds, the latter at an adjusted odds ratio of 2.92 (1.45 to 5.88).

Roughly one in ten tattooed people report an adverse skin reaction. That statement carries two independent samples and is the sturdiest claim on this page.

The 67.5% figure, and why it is not a complication rate

Klügl I and colleagues surveyed 3,411 tattooed participants in German-speaking countries, Dermatology 2010;221(1):43 to 50 (PMID 20215724), and reported that “After tattooing, the people described skin problems (67.5%) or systemic reactions (6.6%). Four weeks after tattooing, 9% still had health problems. Six percent reported persistent health problems.”

The 67.5% counts any skin problem in the immediate post-tattoo window, which substantially captures ordinary healing: redness, swelling, itching, crusting. The comparable figures from that same study are 9% at four weeks and 6% persistent, and those are the ones that sit alongside Brady and Friis. It was a self-selected internet survey, so treat it as a convenience sample.

Four clinic shares, and what each one is really the share of

The Copenhagen tattoo clinic series, Serup J, Sepehri M, Hutton Carlsen K, Dermatology 2016;232(6):668 to 678 (PMID 27974717), classified 493 adverse events in 405 patients seen between 2008 and 2015. Every percentage in it divides by 493 complications, not by tattooed people.

The same series is worth quoting for something it did not find. Verbatim: “We found no cases of cutaneous or other malignancies.” Also in it: 13% papulo-nodular reactions, mainly in black tattoos and considered non-allergic, caused by pigment agglomeration; 9% psycho-social complications.

The red pigment question, answered as precisely as the evidence allows

Among patients who present to dermatology with a tattoo reaction, red and its shades predominate. That is well supported by two clinic series: Serup and Hutton Carlsen's patch-test study of 90 patients, Contact Dermatitis 2014;71(5):255 to 263 (PMID 25040844), calls red “the most predominant colour associated with skin reactions”, and the 493-event series says allergic reactions were “predominantly observed in red tattoos and nuances of red.”

What is not supported is that red carries a dramatically elevated risk per person who chooses it. Only one study compared the share of reactions against the base rate of exposure, and it is Brady 2015. Verbatim: “Forty-four per cent of colour-specific reactions were to red ink, which was only slightly higher than the frequency of red ink in the sampled population (36%). Twenty-five per cent of chronic reactions were to black ink, which was less than expected based on the number of respondents with black tattoos (90.3%).”

Two things follow. The authors' own verdict on red is “only slightly higher”, and with 18 chronic colour-specific reactions in the sample, 44% is on the order of eight people. The genuinely striking result in that sentence is about black: 25% of chronic reactions against 90.3% exposure means black is dramatically under-represented, which is a stronger and better-evidenced signal than the red claim it usually gets buried under.

At population level, both large surveys agree that coloured inks react more than black: Klügl found coloured tattoos provoked more short-term skin reactions (p = 0.003) and systemic reactions (p = 0.0001) than black, and Friis found colours other than black carried higher odds. No study anywhere reports the rate of reaction among people tattooed in red. That denominator does not exist in the accessible literature.

Patch testing does not settle it either. Serup and Hutton Carlsen concluded that “The allergen or allergens responsible for tattoo reactions are not present directly in tattoo ink stock products” and proposed that the allergen results from slow haptenisation in the skin, possibly including photochemical cleavage of red azo pigment. The 2023 review of 25 years of patch-tested tattoo patients states outright that “tattoo allergy cannot be reliably diagnosed via patch testing with today's knowledge.”

Granulomatous and sarcoidal reactions have never been given a denominator

This is a finding rather than a gap in the search. Sarcoidal granulomas on tattoos have been documented for seventy years and can be the first, sometimes the only, cutaneous manifestation of systemic sarcoidosis. Kluger's review in JEADV 2018;32(11):1852 to 1861 (PMID 29763518) is a review of case reports with no denominator. The only percentage in the literature is the 5% clinic share above. Their frequency has never been measured against a population.

The defining investigation is Kennedy BS and colleagues, “Outbreak of Mycobacterium chelonae infection associated with tattoo ink,” New England Journal of Medicine 2012;367(11):1020 to 1024 (PMID 22913660). Nineteen cases arose among the 167 patrons of a single artist between October and December 2011. Pulsed-field gel electrophoresis showed indistinguishable patterns in 11 clinical isolates and in one of three unopened bottles of premixed grey ink. The organism was not isolated from the shop's water or faucet.

The contamination was inside sealed, factory-premixed ink. Not technique, not tap-water dilution, in this outbreak. The finding led to a manufacturer recall. The CDC's MMWR report of 24 August 2012 covers clusters in New York, Washington, Iowa and Colorado, notes contamination found in inks used in two of five clusters and four different brands involved, and adds a line that remains true: “No specific FDA regulatory requirement explicitly provides that tattoo inks must be sterile.” The MMWR incorporates the same New York cluster, so treat these as one investigation reported twice rather than two independent confirmations.

The counterweight is a negative surveillance result and it belongs beside the outbreak. Kotzen M and colleagues used New York City syndromic surveillance to look for tattoo-related mycobacterial infection, PLoS One 2015;10(6):e0130468 (PMID 26076006). Of 31 tattoo-related emergency visits identified, 14 were interviewed, one met the case definition, and that person tested negative. No tattoo-associated cases were reported by city laboratories. Verbatim: “The results were reassuring that an outbreak of NTM in tattoo recipients was not occurring.”

Blood-borne viruses, and the venue distinction that matters most

The two meta-analyses independently agree on the size of the association, a decade apart, with overlapping confidence intervals. They are associations, not transmission rates, and they pool studies that mostly do not stratify by venue while being heavily confounded by injection drug use and incarceration. The venue stratification comes from the CDC's own authors and is the single most important line in this subsection.

Callaghan MF and colleagues, “Safety of Tattoos in Persons Undergoing MRI,” New England Journal of Medicine 2019;380(5):495 to 496 (PMID 30699316). Between 2011 and 2017, 330 persons aged 18 to 66 underwent 3 tesla imaging across 585 sessions, covering 932 unique tattoos, 717 of them black. Specific absorption rate was held below 2 watts per kilogram.

Two events occurred. One participant retrospectively reported awareness and tingling, which the investigators did not classify as a tattoo-related reaction. One participant reported a warm, tight feeling around a wrist tattoo during the localiser sequence; the scan was stopped and the sensation resolved fully within 24 hours without intervention. That single classified event produces two different rates with two different denominators: 0.17% per scan session (1 of 585) and 0.30% per person (1 of 330, 95% CI 0.01 to 1.68). For a consumer reference page the per-person figure is the right one, and it must be labelled as such.

The study's own inclusion criteria bound the result and must travel with it. Participants needed no more than 5% of the body tattooed, no tattoo longer than 20 cm, and no tattoos on the head, neck or genitals. Large-scale bodysuits and neck work were not tested. The authors' hedge is “under these specific study conditions.”

Two independent sources converge. The Rhineland Study scanned 3,639 participants at 3 tesla, of whom 305 had tattoos or permanent make-up, and reported in Frontiers in Neurology 2022;13:795573 (PMID 35392639) that “None of the participants reported adverse events.” Tope and Shellock surveyed 1,032 people with permanent cosmetics in J Magn Reson Imaging 2002;15(2):180 to 184 (PMID 11836774); 135 had actually been scanned, and 2 of those 135 (1.5%) had transient tingling or burning. That 1.5% divides by 135, not by 1,032, and anyone quoting it against the larger number has swapped denominators.

Guidance is screening, not exclusion. The 2019 American College of Radiology guidance document notes that “conductive loops may be created by skin adornments such as tattoos, especially with dark colors of ink (black, brown, and blue) and curved patterns” and recommends continued vigilance when screening. There is no contraindication to scanning tattooed patients. The FDA's own position, verbatim: “the risks of avoiding an MRI when your doctor has recommended one are likely to be much greater than the risks of complications from an interaction between the MRI and tattoo or permanent makeup. Instead of avoiding an MRI, individuals who have tattoos or permanent makeup should inform the radiologist or radiologic technologist.”

A related belief, that a tattoo prevents an epidural, has been retired in the literature. Kluger and Sleth, Presse Med 2020;49(4):104050: “To date, no convincing complication has been ever reported after an EA through a tattoo.”

Following the citation chain

The chain can be traced because one open-access paper states the figure and cites it. Fauger A and colleagues, “Tattoo aftercare management with a dermo-cosmetic product,” J Cosmet Dermatol 2022;21(3):1051 to 1056 (PMID 33884740), writes: “Tattoo healing is usually complete after 2-3 weeks”, with two references.

The only objective in-vivo observation has a sample of one

Kröger M and colleagues, Dermatology 2023;239(3):478 to 493 (PMID 36787702), used two-photon excited fluorescence lifetime imaging in living skin, down to the reticular dermis. Sample, verbatim: “One subject with a freshly applied tattoo and 10 subjects with tattoos applied over 3 years ago.” One person for the fresh tattoo.

The key result, verbatim: “The carbon black particles deposited around the incision have still been visible 84 days after tattoo application, showing delayed recovery of the epidermis.” The authors also observed that collagen I in tattooed skin showed directionality similar to scar tissue, with greater firmness and reduced elasticity.

Weight this honestly. One subject is not a healing-duration study, and no replication was found. But it is the only objective in-vivo time course of a fresh tattoo located anywhere in the literature, and at 84 days it describes epidermal recovery as still incomplete. It points to healing taking longer than the conventional figure, not shorter. Two to three weeks may fairly describe when a tattoo looks and feels healed. It is not a claim about tissue.

The literature says outright that the guidance is not evidence-based

Liszewski W and colleagues, “An Analysis of the Content and Recommendations of 700 American Tattoo Aftercare Instructions,” Dermatology 2023;239(6):988 to 995 (PMID 37604151), collected 700 aftercare instructions from all 50 states and Washington D.C. Verbatim:

Only 49.9% of the instructions gave any parameter for when to contact the tattooist, and only 19.4% for when to contact a physician. Seventy different moisturisers were recommended across the set, 22 of them niche tattoo-specific products, and 14.9% encouraged topical antibiotics. With 700 documents behind it, this is the strongest available citation for the aftercare evidence gap, and it names the mechanism of the gap explicitly.

Three attributions to avoid

What has been measured, and what it is not

• Barrier function in healed tattoos. Nørreslet LB and colleagues, Skin Res Technol 2019;25(3):382 to 388, measured 28 tattoos in 26 people and found no significant difference in transepidermal water loss between tattooed and non-tattooed skin (medians 6.6 vs 7.2 g/m²/h). Cross-sectional, on already-healed tattoos. Their opening line concedes the point: “Initially after tattooing, the skin barrier function is broken. However, the long-term impact of clinically healed tattoos on this has never been studied.”
• A skin model, not a person. Reddersen K and colleagues, Scientific Reports 2025;15(1):2277, tattooed a three-dimensional human skin model and tracked it for 7 days: a strong inflammatory reaction that “subsided 4 days after treatment”, with the healing phase detectable in gene expression. It is a laboratory model.
• Visible inflammation at 14 days. Fauger 2022 observed 30 people for 14 days and reported redness absent in 100% and 96% of subjects by artist and self-assessment, with oedema completely gone. It is described by its own authors as a survey, has no control arm, the objective assessment was performed by the tattoo artist, three of five authors are employees of the product manufacturer, and the window was fixed at 14 days so it could not have detected healing that took longer. It establishes that visible inflammation is gone by day 14 in 30 people. Nothing more.
• A legitimate analogy, labelled as one. Partial-thickness wound re-epithelialisation has been measured. A randomised trial in 40 patients over 65, Eur Burn J 2024;5(4):335 to 345, reported mean time to healing of 31.7 and 27.3 days for split-thickness donor sites. That is an elderly cohort and a far deeper, larger wound than a tattoo. A tattoo is a repeated micro-puncture injury depositing pigment into the papillary and reticular dermis, closer in character to microneedling than to a donor site. Converting a donor-site number into a tattoo healing time would be exactly the error this page exists to prevent.

The search that produced the negative

Twelve queries were run against the Europe PMC REST API and PubMed E-utilities, and every hit set was screened. Query strings included title-restricted searches on tattoo with wound healing and re-epithelialisation (64 hits), tattoo with transepidermal water loss (62), tattoo with aftercare (35), tattoo with healing time (80), tattooing with prospective and healing (246), tattoo with epidermal or dermal healing or wound closure (153), and tattoo with clinical trial or randomised and healing (26), plus two independent PubMed cross-checks. No study measuring tattoo healing duration by objective criteria in a human cohort was found in any of them. The only objective in-vivo observation returned anywhere was Kröger 2023, at one subject.

How strong is the evidence base

The current best synthesis is a 2026 systematic review of Nd:YAG lasers in J Cosmet Dermatol 25(6), registered as PROSPERO CRD420251241094, covering 46 studies. Three of its own statements set the ceiling on every claim below. Verbatim: quantitative meta-analysis “was not feasible” given the heterogeneity, so findings were synthesised narratively; using GRADE, “the overall certainty of evidence … was rated as low to very low”; and “Nd:YAG lasers achieved meaningful clearance of black and blue tattoo pigments in most studies, generally requiring 4-8 sessions to reach ≥ 75% clearance.”

The randomised base underneath it is tiny. A 2022 systematic review in J Am Acad Dermatol 87(1):103 to 109 screened 3,037 studies and included 36: seven randomised controlled trials, two non-randomised controlled trials, and 27 case series. A 2021 review of high-power Q-switched Nd:YAG identified 122 articles and included six, covering 188 individuals. There is no Cochrane systematic review of tattoo removal; a Europe PMC query restricted to the Cochrane database returned zero hits and the 2026 review searched the Cochrane Library and included none.

Number of sessions: what actually exists

There is one published prediction tool. Kirby W and colleagues, “The Kirby-Desai Scale,” J Clin Aesthet Dermatol 2009;2(3):32 to 37 (PMID 20729941), is a retrospective chart review of 100 clinic patients scoring six parameters: Fitzpatrick skin type, location, colour, amount of ink, scarring and layering. It reported a correlation coefficient of 0.757 and, verbatim, “The average number of treatments required was 10 treatments (9.91±3.18), with a range of 3 to 20 treatments.”

The paper also reports complete removal in 100% of its 100 patients, and that figure is a selection artefact the authors describe themselves: patients who “did not complete the expected tattoo-removal treatment sessions” because of pain, missed appointments or loss to follow-up were excluded by design. The paper never defines what “satisfactory removal” means; the full text was searched and no operational definition exists.

Three groups in three countries, using different lasers and protocols, all found actual session counts below prediction. What the evidence supports is this: black and blue pigment generally needs 4 to 8 sessions to reach at least 75% clearance across 46 studies of low to very low certainty; the only prediction tool is a single-clinic retrospective heuristic whose cohort averaged about 10 sessions with a range of 3 to 20; recent prospective work reports averages around 6 in 116 patients. Fixed ranges such as “five to ten sessions” circulate widely in removal marketing. No primary source in this literature supports one, and no validated per-colour or per-skin-type session table exists at all.

On skin type specifically, the Kirby-Desai scale weights Fitzpatrick type most heavily of its six parameters, and the largest prospective dataset contradicts that: Menozzi-Smarrito and Pineau found that ink density mattered most, followed by location, age and design technique, while skin type “showed no significant influence.” That study covered Fitzpatrick types I to IV only, so it says nothing about types V and VI where melanin competition is greatest. Fitzpatrick type is well documented as a complication risk factor, which is a different claim from session count and is routinely conflated with it.

Complete clearance: the number everyone wants

The spread from 1.26% to 100% is not measurement noise. It is entirely explained by whether the denominator includes the people who stopped coming. The most defensible consumer-facing figure is Klein's 38% of 157, because it is the only one drawn from a patient-side sample rather than a clinic's completed-course records; its weakness is that it is self-reported and self-selected. The FDA's own summary is blunt: “Complete removal without scarring may be impossible.” Kilmer added a caution in 1993 that still stands: “Histopathologic examination demonstrated persistence of tattoo ink in clinically clear areas.” Clinical clearance is not histological clearance.

Complications: two credible sources, two orders of magnitude apart

Poelhekken M and colleagues surveyed 173 practitioners, 94 of them actively performing laser removal, in Dermatology 2026;242(4):449 to 456 (PMID 41980006). These are practitioner estimates per treatment session, not audited counts. Verbatim: complications in “an estimated 8.1% of laser tattoo removal treatment sessions, most commonly oedema (3.0%), blistering (1.9%), and haematoma (0.7%)”, with hypo- and hyperpigmentation around 0.6%, hypertrophic scarring and paradoxical darkening at 0.2%, and keloid, allergic reaction, ink blow-out and infection each at 0.1%.

Klein A and colleagues surveyed 157 patients in German-speaking countries, Lasers Med Sci 2014;29(2):729 to 738 (PMID 23907603). Verbatim: “Local transient side effects occurred in nearly all participants, but an important rate of slightly visible scars (24%) or even important scarring (8%) was reported”, and “nearly half of the participants experienced hypopigmentation in the treated area.”

Practitioner-estimated hypopigmentation at roughly 0.6% per session against patient-reported hypopigmentation at roughly half per course. Some of that gap is per-session against per-course, some is who is doing the reporting, and the studies are twelve years apart with different laser generations. Neither can be dismissed. If a page quotes one, it has to quote both and name the denominator basis.

The mechanism behind hypopigmentation makes it a physics problem rather than a technique problem. The JRC states that at shorter wavelengths “the epidermal melanocytes are destroyed preferentially”, that operating Nd:YAG at 1064 nm minimises the risk, but that removing red, yellow and orange requires the 532 nm wavelength, “making hypopigmentation unavoidable.” Permanent hypopigmentation “could affect up to 10% of the individuals having removed their tattoos with laser.” A three-laser comparison on 42 blue-black tattoos found the Q-switched ruby laser had the highest clearance and the highest incidence of long-lasting hypopigmentation, while the Nd:YAG had none. For darker skin, removing warm colours and avoiding hypopigmentation are in direct physical conflict.

Paradoxical darkening is rare overall and common in specific inks. The foundational description is Anderson RR and colleagues, Arch Dermatol 1993;129(8):1010 to 1014 (PMID 8352605), five cases of “immediate, irreversible darkening of cosmetic, white, flesh (skin-color), and pink-red colored tattoos”, where further laser treatment failed in two cases and surgical excision was necessary. The probable mechanism is reduction of ferric oxide to ferrous oxide, and the authors state the chemical reaction “remains unknown.” Their recommendation was test-site exposure before treatment. Rates depend entirely on which population is meant: 0.2% of all sessions in the practitioner survey; 30% of 33 patients in a retrospective review of facial cosmetic tattoo removal (PMID 36946770); 33 of 184 non-black tattoos in the JRC's figures.

Picosecond against nanosecond: unresolved, except on pain

The 2026 systematic review adjudicates: picosecond superiority over conventional Q-switched systems “remains unproven and cannot be established with certainty given the current paucity of evidence.” Design differences explain much of the divergence, since Pinto allowed only two sessions and Bäumler's comparator was a ruby laser rather than Nd:YAG. One picosecond advantage is consistent across trials, and it is pain, not clearance: Pinto measured 3.8 against 7.9 on the pain scale (P < 0.001), and Bäumler found significantly less pain with picosecond pulses.

Which colours resist, and why

A pigment clears only if it absorbs the laser's wavelength. Black absorbs broadly, so almost any wavelength works on it. Red absorbs green light at 532 nm. Green absorbs red light at 694 nm from a ruby laser or 755 nm from an alexandrite. Yellow and white absorb very little across the window these lasers operate in, which is why they resist. Kilmer's 1993 series of 25 patients and 39 tattoos found an excellent response, defined as more than 75% ink removal, in 77% of black tattoos, and stated plainly that “Colored inks were not as effectively removed.” Lorgeou's 49-patient study found that only one of its five polychromic tattoos reached 75% improvement with either laser type.

Pigment breakdown products: a demonstrated mechanism with an unquantified human dose

Every study in this area is in vitro, on ink suspensions or purified pigment. None has measured human in-vivo exposure. Vasold R and colleagues irradiated two azo compounds in suspension in 2004 and found cleavage products including 2-methyl-5-nitroaniline and 4-nitro-toluene, while also reporting that “the tattoo colorants already contain such compounds before laser irradiation”; the authors themselves framed the human risk assessment as still to be done. Schreiver I and colleagues, Scientific Reports 2015;5:12915 (PMID 26243473), identified hydrogen cyanide, benzene and benzonitrile among fragmentation products of ruby laser irradiation of copper phthalocyanine blue in suspension. That finding is regularly and wrongly attributed to carbon black. It was the blue pigment.

The framing sentence comes from the systematic review of the whole area, Fraser TR and colleagues, J Expo Sci Environ Epidemiol 2022;32(3):343 to 355 (PMID 34274958), verbatim: the review “demonstrates that there is a lack of knowledge regarding tattoo pigment degradation/metabolism, with only eleven articles found pertaining to the photolysis of tattoo pigments and two articles on the metabolism of tattoo pigments. The limited research indicates that the photolysis of tattoo pigments could result in many toxic degradation products.” Eleven papers and two papers, and the review's own verb is “could”. Any claim that laser removal does expose people to carcinogens at harmful doses goes beyond what has been shown.

On regret, for context rather than as a removal argument: Pew Research Center, surveying 8,480 US adults between 10 and 16 July 2023, found that 32% of Americans have a tattoo, 22% have more than one, and “about a quarter (24%) say they ever regret getting one or more of their tattoos.” Published European regret figures range from 5% to 50%, which the JRC prefaces with “Although data are scarce.” That spread is a measurement failure, not a finding.

1. No clinical study has measured how long a tattoo takes to heal. Twelve database queries, all screened. The universal two-to-three-week figure traces to a 2012 narrative review and a 1992 forensic-pathology review.
2. The American Academy of Dermatology publishes no tattoo healing duration. Its page was fetched and read. Attribute no timeline to it, or to any other named body.
3. IARC has never evaluated tattooing or permanent make-up. Its own 2024 Advisory Group says so verbatim and rates it medium priority.
4. The IARC carbon black monograph has nothing to say about tattooing. The word appears zero times across all 466 pages.
5. Benzo[a]pyrene has never been detected in tattooed human tissue. It was below the detection limit in all 32 specimens in the only study that looked for it.
6. No study reports the rate of allergic reaction among people tattooed in red ink. Only the share of reactions involving red, which is a different quantity.
7. No population rate exists for granulomatous or sarcoidal tattoo reactions. The only percentage in the literature is a share of clinic complications.
8. No pooled complete-clearance rate for laser removal exists. The 46-study systematic review found meta-analysis not feasible and rated certainty low to very low.
9. No generalisable session-count figure for laser removal exists, and no session data stratified by Fitzpatrick skin type exists at all.
10. There is no Cochrane systematic review of tattoo removal.
11. No contamination rate for tattoo ink can be computed from FDA recall data. No published denominator of units sold exists.
12. California sets no chemical composition limits for tattoo ink. Section 119311(b) is a provenance rule, not a composition rule, and there is no US analogue to REACH entry 75.

A thirteenth belongs here for a different reason: no colour additive is FDA-approved for injection into skin. That is an absence of approval rather than an absence of authority, and it is the reason California requires every tattoo client in the state to be told so in writing.

Regulation and institutional documents

• Commission Regulation (EU) 2020/2081 of 14 December 2020 amending Annex XVII to Regulation (EC) No 1907/2006 (REACH) as regards substances in tattoo inks or permanent make-up. EU-adopted text. EUR-Lex deep links redirected to the Official Journal homepage in every form tried, and ECHA returned a firewall block, so this republication of the adopted text was used.
• German Federal Institute for Risk Assessment (BfR), Frequently asked questions about tattoo inks, last updated 5 June 2025. Includes BfR Opinion No. 039/2020.
• United States Food and Drug Administration, “Tattoos & Permanent Makeup: Fact Sheet” and “Tattoos, Temporary Tattoos & Permanent Makeup”, content current as of 15 October 2024, fda.gov. FDA Constituent Update, 24 October 2024. FDA recall notice for Sierra Stain LLC, 23 August 2024.
• Modernization of Cosmetics Regulation Act of 2022, Public Law 117-328, division FF, title III, section 3502; 21 U.S.C. section 364h, text current as of 28 August 2026, uscode.house.gov.
• California Health and Safety Code, division 104, part 15, chapter 7 (sections 119300 to 119328), California's Safe Body Art Act, and California Penal Code sections 652 and 653, leginfo.legislature.ca.gov.
• Los Angeles County Department of Public Health, Environmental Health, Body Art Program.
• European Commission Joint Research Centre, Safety of tattoos and permanent make-up: Final report, EUR 27947 EN, 2016. Companion reports EUR 27528 EN (2015) and EUR 27672 EN (2016).
• IARC, Monographs Volume 93: Carbon Black, Titanium Dioxide, and Talc, Lyon, 2010. Full PDF downloaded and searched.
• IARC, Report of the Advisory Group to Recommend Priorities for the IARC Monographs during 2025-2029, November 2024, agent 159.
• American Academy of Dermatology, Caring for tattooed skin. Cited as a verified negative: it publishes no healing duration.
• Pew Research Center, 32% of Americans have a tattoo, 15 August 2023, 8,480 US adults.
• Kantonales Laboratorium Basel-Stadt tattoo ink survey, 2014, 229 inks. EN 17169:2020, Tattooing: safe and hygienic practice (CEN), paywalled and quoted only via Fauger 2022. Vegan Society Trademark standards and Vegan Action certification.

Ink composition, metals and translocation

• Moseman K, et al. Analysis of tattoo inks by Raman, XRF and NMR. Anal Chem 2024;96(9):3906-3913. PMID 38387033. 54 inks, 9 brands.
• Moseman K, et al. Analysis of blue and green REACH compliant tattoo inks. Analyst 2024;149(21):5329-5335. PMID 39327926. 10 inks, 5 manufacturers.
• Serup J, et al. Danish tattoo ink registry. Dermatology 2025;241(3):259-271. 10,833 bottles by label. Raw counts only; the percentage table is internally inconsistent.
• Schreiver I, et al. Pyrolysis of purified tattoo pigments. Arch Toxicol 2016;90(7):1639-1650. 36 pigments.
• Heavy metals in EU-market tattoo inks. Toxics 2025;13(11):934, 30 October 2025. 41 products.
• Violi F, et al. J Hazard Mater 2026;502:140874. PMID 41554661. 15 Australian inks, abstract level only, ScienceDirect blocked.
• Wang X, et al. Contact Dermatitis 2021;85(3):340-353. PMID 34089526. 73 inks.
• Regensburger J, et al. Experimental Dermatology 2010;19(8):e275-281. PMID 20545755. 19 inks. Frequently miscited to Archives of Dermatological Research; it is Experimental Dermatology.
• Lehner K, et al. PLoS One 2014;9(3):e92787. 16 tattooed skin specimens and 16 matched lymph nodes.
• Schreiver I, et al. Sci Rep 2017;7(1):11395. PMID 28900193. 4 tattooed cadaver donors and 2 controls.
• Schreiver I, et al. Tattoo needle wear particles. Part Fibre Toxicol 2019;16(1):33. PMID 31451117.
• Maarouf M, et al. J Clin Aesthet Dermatol 2019;12(2):37-38. The single literature hit for “vegan tattoo”, an editorial that leans cautionary.

Cancer

• Nielsen C, Jerkeman M, Jöud AS. eClinicalMedicine 2024;72:102649. PMID 38827888.
• Clemmensen SB, et al. BMC Public Health 2025;25(1):170. PMID 39819495.
• Warner FM, et al. Cancer Epidemiol Biomarkers Prev 2020;29(10):2093-2095. PMID 32699076.
• McCarty RD, et al. Cancer Med 2024;13(20):e70260. PMID 39444249.
• McCarty RD, et al. Tattooing and risk of melanoma. J Natl Cancer Inst 2025;117(12):2495-2504. PMID 40839395.
• McConnell T, et al. eClinicalMedicine 2025;89:103563. PMID 41140458. PROSPERO CRD42024586505.
• Tudella GCN, et al. Clin Transl Oncol, online 6 June 2026. PMID 42250187.
• Kluger N, Koljonen V. Tattoos, inks, and cancer. Lancet Oncol 2012;13(4):e161-168. PMID 22469126.

Complications, infection and imaging

• Brady BG, Gold H, Leger EA, Leger MC. Contact Dermatitis 2015;73(2):91-99. PMID 26016445. 300 participants. Frequently misattributed to other authors.
• Friis K, et al. Dermatology 2024;240(2):297-303. PMID 38081147. 5,914 tattooed respondents.
• Klügl I, et al. Dermatology 2010;221(1):43-50. PMID 20215724. 3,411 participants, self-selected internet survey.
• Serup J, Sepehri M, Hutton Carlsen K. Dermatology 2016;232(6):668-678. PMID 27974717. 493 adverse events in 405 patients.
• Serup J, Hutton Carlsen K. Contact Dermatitis 2014;71(5):255-263. PMID 25040844. 90 patch-tested patients.
• Schubert S, Kluger N, Schreiver I. Contact Dermatitis 2023;88(5):331-350. PMID 36772861.
• Kluger N. Tattoo-associated uveitis. JEADV 2018;32(11):1852-1861. PMID 29763518.
• Kennedy BS, et al. N Engl J Med 2012;367(11):1020-1024. PMID 22913660. CDC, MMWR 2012;61(33):653-656.
• Kotzen M, et al. PLoS One 2015;10(6):e0130468. PMID 26076006.
• Jafari S, et al. Int J Infect Dis 2010;14(11):e928-940. PMID 20678951. Lim SH, et al. PLoS One 2022;17(1):e0262990. PMID 35085358. Tohme RA, Holmberg SD. Clin Infect Dis 2012;54(8):1167-1178. PMID 22291098.
• Callaghan MF, et al. N Engl J Med 2019;380(5):495-496. PMID 30699316. Published PDF read directly; a circulating summary giving 573 sessions on 319 individuals is wrong.
• Lohner V, et al. Front Neurol 2022;13:795573. PMID 35392639. Tope WD, Shellock FG. J Magn Reson Imaging 2002;15(2):180-184. PMID 11836774. Greenberg TD, et al. ACR guidance document on MR safe practices, 2019. Kluger N, Sleth JC. Presse Med 2020;49(4):104050. PMID 32768613.

Healing

• Kröger M, et al. Dermatology 2023;239(3):478-493. PMID 36787702. One subject with a fresh tattoo.
• Fauger A, et al. J Cosmet Dermatol 2022;21(3):1051-1056. PMID 33884740. Its references 4 and 5: Kluger N, Am J Emerg Med 2012;30(9):2055-2063 (PMID 22944541) and Sperry K, Am J Forensic Med Pathol 1992;13(1):7-17 (PMID 1585890). Sperry's full text was not accessible; only its abstract was read.
• Liszewski W, et al. 700 American tattoo aftercare instructions. Dermatology 2023;239(6):988-995. PMID 37604151. Earlier viewpoint: Liszewski W, Jagdeo J, Laumann AE. JAMA Dermatol 2016;152(2):141-142. PMID 26535490.
• Nørreslet LB, et al. Skin Res Technol 2019;25(3):382-388. PMID 30600550. Serrano-Serra JP, et al. J Clin Med 2021;10(4):888. PMID 33671713. Reddersen K, et al. Sci Rep 2025;15(1):2277. PMID 39833528.
• Cussons D, et al. Eur Burn J 2024;5(4):335-345. PMID 39727907. Cited as a labelled analogy only.

Laser removal

• Efficacy of Nd:YAG lasers for tattoo removal: systematic review. J Cosmet Dermatol 2026;25(6). PROSPERO CRD420251241094. 46 studies, GRADE low to very low.
• Gurnani P, et al. J Am Acad Dermatol 2022;87(1):103-109. PMID 32763326. Modena DAO, et al. J Cosmet Laser Ther 2021;23(3-4):41-48. PMID 34668826. Reiter O, et al. Lasers Med Sci 2016;31(7):1397-1405. PMID 27311768.
• Kirby W, Desai A, Desai T, Kartono F, Geeta P. J Clin Aesthet Dermatol 2009;2(3):32-37. PMID 20729941. 100 completers.
• Aurangabadkar SJ, et al. J Cutan Aesthet Surg 2019;12(2):95-104. PMID 31413477. Egozi E, Toledano O. J Cosmet Dermatol 2024;23(3):818-823. PMID 38308455. Menozzi-Smarrito C, Pineau N. J Cosmet Dermatol 2025;24(7):e70186. PMID 40682360.
• Ross V, et al. Arch Dermatol 1998;134(2):167-171. PMID 9487208. Pinto F, et al. Br J Dermatol 2017;176(2):457-464. PMID 27518129. Lorgeou A, et al. JEADV 2018;32(2):265-270. PMID 28758261. Bäumler W, et al. JEADV 2022;36(2):305-312. PMID 34543473.
• Kilmer SL, et al. Arch Dermatol 1993;129(8):971-978. PMID 8352621. Anderson RR, et al. Arch Dermatol 1993;129(8):1010-1014. PMID 8352605. Leuenberger ML, et al. Dermatol Surg 1999;25(1):10-14. PMID 9935085.
• Poelhekken M, et al. Dermatology 2026;242(4):449-456. PMID 41980006. Klein A, et al. Lasers Med Sci 2014;29(2):729-738. PMID 23907603. Hartman N, et al. Dermatol Surg 2023;49(6):559-565. PMID 36946770.
• Vasold R, et al. Photochem Photobiol 2004;80(2):185-190. PMID 15244509. Schreiver I, et al. Sci Rep 2015;5:12915. PMID 26243473. Fraser TR, et al. J Expo Sci Environ Epidemiol 2022;32(3):343-355. PMID 34274958.

Blocked or paywalled during research, and noted rather than worked around silently: EUR-Lex deep links, ECHA, the Cochrane Library interface, ScienceDirect, Wiley, Karger, and the Journal of Investigative Dermatology. Figures reachable only through a systematic review's extraction table are labelled as such where they appear.