Does Sunscreen Cause Cancer? What the Only Trial Actually Found

“Sunscreen causes cancer” trended on social media again this summer. Does sunscreen cause cancer? The claim has two legs, and both rest on things that actually happened. Sunscreen filters really are absorbed into the bloodstream; regulators confirmed it. Skin cancer rates really did climb through decades of rising sunscreen use; cancer registries record it. What follows is what the primary sources say when you read past those two facts.

The two claims, separated

The first claim is about absorption. FDA studies published in 2019 and 2020 found that several organic UV filters pass through skin into blood plasma, and we went through the plasma concentrations and the regulatory thresholds in our breakdown of the sunscreen absorption study. What those results do not establish is harm. The American Academy of Dermatology put the distinction plainly: “Keep in mind what the researchers pointed out. Just because an ingredient is absorbed into the bloodstream does not mean that it is harmful or unsafe.”^[15]

The second claim is the harder one. Melanoma incidence in fair-skinned populations rose for decades while sunscreen use rose alongside it, and some observational studies found more melanoma among sunscreen users than among non-users. Those numbers exist. The rest of this article is about what produced them.

The only randomized trial of sunscreen and melanoma

Between 1992 and 1996, researchers in Nambour, a town in subtropical Queensland, randomly assigned 1,621 adults aged 25 to 75 to one of four arms in a 2×2 factorial design: daily SPF 15+ sunscreen or discretionary use, crossed with 30 mg of beta-carotene or placebo.^[1] The daily group applied sunscreen to head, neck, arms, and hands for four and a half years, and participants were followed for more than a decade afterward.

Randomization is what separates this trial from the rest of the literature. An observational study compares people who chose sunscreen against people who did not, and those groups differ in dozens of ways. Here, a coin flip decided.

Outcome	Result	Source
Squamous cell carcinoma, trial period	By tumor count: 1,115 vs 1,832 per 100,000; RR 0.61 (0.46-0.81). By first new cancer per person: 876 vs 996 per 100,000; RR 0.88 (0.50-1.56), not significant	Green, Lancet 1999 [1]
Squamous cell carcinoma, including 8 further years of follow-up	“significantly decreased by almost 40%”; rate ratio 0.62 (0.38-0.99)	van der Pols, CEBP 2006 [2]
Invasive melanoma, ten years after trial cessation	3 vs 11 cases; HR 0.27 (0.08-0.97)	Green, JCO 2011 [3]
All melanoma, same follow-up	11 vs 22 cases; HR 0.50 (0.24-1.02), P = .051	Green, JCO 2011 [3]
Basal cell carcinoma	No significant reduction at either point	Green 1999 [1]; van der Pols 2006 [2]
Solar keratoses	24% fewer	Darlington, Arch Dermatol 2003 [4]
Photoaging	24% less (relative odds 0.76, 0.59-0.98)	Hughes, Ann Intern Med 2013 [5]

Two rows in that table are the ones most often reported wrong.

The melanoma result is narrower than the headlines. All new melanomas came to 11 in the daily group against 22 in the discretionary group: hazard ratio 0.50, confidence interval 0.24 to 1.02, P = .051.^[3] That interval includes 1, and .051 does not clear the conventional threshold. What reached significance was the invasive subgroup, 3 cases against 11, HR 0.27 (0.08-0.97). A hazard ratio of 0.27 works out to roughly 73% lower observed incidence, but an upper bound of 0.97 leaves the effect size loosely bounded. “Sunscreen halved melanoma” is not what this paper reports.

Basal cell carcinoma was not prevented. Not during the trial, and not over the extended follow-up, where it “tended to decrease but not significantly.”^[1][2] Any claim that sunscreen prevents all skin cancer runs into this trial’s own data. What was reduced is specific: squamous cell carcinoma by tumor count (though not by the number of people developing a first cancer), invasive melanoma, precancerous keratoses, photoaging. Basal cell carcinoma is not on the list.

What the trial did not find is also part of the record. The Lancet report: “There was no harmful effect of daily use of sunscreen in this medium-term study.”^[1]

The objection this trial already answers

A common response to any sunscreen study is that it does not apply to real life, because nobody applies enough. Lab SPF ratings are measured at 2 mg/cm², and surveys find people using a fraction of that. If the Nambour participants had been coached into a full lab dose, the argument goes, the results say nothing about your morning routine.

Someone measured. A 2002 analysis of the same trial reported a median application density of 0.79 mg/cm², “less than half the amount needed to achieve the labeled sun protection factor.”^[6] Fifty-six percent applied on five or more days per week.

That number matters because of what a later study did with it. In 2018, researchers testing DNA damage in human skin used 0.75 mg/cm² as their “typical use” condition, against 1.3 and 2.0 mg/cm².^[7] At a median of 0.79, the Nambour participants were applying almost exactly what photobiology researchers treat as the sloppy real-world baseline.

So the reductions in the table above are not what sunscreen achieves under ideal conditions. They are what it achieved among people applying it the way people actually do. The 2018 study also showed thicker layers doing measurably better: cyclobutane pyrimidine dimer formation dropped significantly at 1.3 and 2.0 mg/cm², while the 0.75 condition reached significance only under repeated five-day exposure.^[7] More product still buys more protection, which is a separate question from whether a normal amount does anything. How lab SPF numbers translate to real application is covered in our piece on whether you can trust Korean SPF labels.

What the large observational studies show

The biggest cohort study followed 143,844 Norwegian women aged 40 to 75 across 1,532,247 person-years, recording 722 melanomas.^[8] Women using SPF 15 or higher had a hazard ratio of 0.67 (0.53-0.83) against women using SPF below 15. If every woman in that age range used SPF 15+, the authors estimated, the population could “potentially reduce their melanoma incidence by 18%.”

Read that comparison carefully, because it is routinely mis-stated. This is high-SPF users against low-SPF users, not sunscreen users against people who use none. And the same paper flags its own confounding: “Sunscreen users reported significantly more sunburns and sunbathing vacations and were more likely to use indoor tanning devices.”^[8]

That sentence is the key to the whole observational literature. A 2019 meta-analysis pooled 28 studies covering 21,069 melanoma cases. The answer it found depends entirely on study design:^[9]

Study design	Pooled estimate
Hospital-based case-control	OR 0.57 (0.37-0.87)
Ecological	RR 0.48
Randomized controlled trial	HR 0.49 (0.24-1.01)
Population-based case-control	OR 1.17 (0.90-1.51)
Cohort	HR 1.27 (1.07-1.51)

The cohort row is where the social media claim comes from, and it is not fabricated. Pooled cohort studies do show a 27% higher melanoma rate among sunscreen users, with a confidence interval that excludes 1. The meta-analysis authors give their own reading: “Evidence from observational studies on sunscreen use and melanoma risk was weak and heterogeneous, consistent with the challenges of controlling for innate confounding by indication. The only RCT showed a protective effect of sunscreen.”^[9]

Confounding by indication is the technical name for a simple problem. People who burn in fifteen minutes buy sunscreen; people who tan without effort often do not. Fair skin drives both the purchase and the melanoma, so the two travel together in any dataset where people chose for themselves. That is why the randomized design carries weight out of proportion to its 1,621 participants.

Why the rates rose anyway

A 2023 review in Cancers took up the question under the title “Possible Explanations for Rising Melanoma Rates Despite Increased Sunscreen Use over the Past Several Decades.”^[12] The authors set out five explanations and are careful to call them hypotheses rather than conclusions:

• Overdiagnosis and expanded screening, as dermoscopy and skin checks became routine
• Risk compensation: sunscreen use extending deliberate time in the sun
• Sunscreens available before the 1990s lacking broad-spectrum UVA protection
• Reactive oxygen species generated by some UV filter compounds
• Changes in ozone and UV index

The third points at a timing problem that undercuts the “sunscreen failed” framing. Melanoma diagnosed in 2020 reflects exposure from the 1970s and 1980s, when modern broad-spectrum sunscreen was not widely available. Comparing this decade’s diagnoses to this decade’s sunscreen sales lines up two numbers from different eras.

If the timing argument holds, the cohorts young enough to have used modern sunscreen their whole lives should bend downward first. In Queensland, they did. Registry data from 1982 to 2010 in people aged 15 to 24 show rising incidence early on, then a reversal: invasive melanoma in males climbed 3.3% a year until 1994 and then fell 4.2% a year (95% CI −6.2 to −2.1); in females it rose 1.7% a year until 1997 and then fell 4.9% a year (−7.6 to −2.2).^[10] The authors summarize: “From 1994 to 2010 in males and from 1997 to 2010 in females, the incidence rates of melanoma in young people aged between 15 and 24 years in Queensland have been in substantial decline by an average 4-5% per year.” Note the shape of that curve: rates went up first and only turned over from the mid- to late 1990s, so descriptions of Australian melanoma rates halving since the 1980s get the early years backwards.

A demographic explanation competes with the sunscreen one: Australia’s population became less European over the same period, and lighter skin carries higher melanoma risk. A 2024 analysis in the Medical Journal of Australia tracked that high-risk ancestry share falling from 85.3% to 71.1% between 2006 and 2021, yet concluded that “Melanoma incidence has declined in some younger age groups in Australia, including among people with high risk ancestry.”^[11] The decline appears inside the high-risk group, not only in the population average. Its annual percent changes below age 34 were “generally not statistically significant,” so it corroborates the Queensland data rather than standing alone.

The benzene question

In May 2021, the independent testing laboratory Valisure filed a citizen petition with the FDA after analyzing 294 batches across 69 brands of sunscreen and after-sun products. Its finding, verbatim: “In summary, 78 product batches had detectable levels of benzene, 26 contained benzene in concentrations between 0.1 ppm and 2.0 ppm and 14 contained over 2 ppm.”^[13] The highest reading, 6.26 ppm, came from a Neutrogena Ultra Sheer Weightless Sunscreen Spray SPF 100+, against an FDA conditional limit of 2 ppm. Benzene is an IARC Group 1 carcinogen, and serious recalls followed.

Company	Class	Initiated	Quantity
Johnson & Johnson Consumer (Neutrogena)	Class I	2021-07-14	1,437,804 cans
Beiersdorf (Coppertone)	Class I	2021-09-30	142,236 cans
Edgewell (Banana Boat)	Class I	2022-07-29	139,680 aluminum spray cans

Three things in the FDA’s own enforcement records cut against the reading that sunscreen chemistry produces benzene.

The agency filed it as contamination, not composition. The recall reason recorded for the Johnson & Johnson action is “Chemical Contamination: presence of benzene.”^[14] Benzene appears on no sunscreen ingredient list. It entered the products from somewhere in manufacturing.

Mineral sunscreens were recalled too. Four of the five Coppertone products in that recall were mineral formulas: the Pure & Simple baby, kids, and adult versions, plus the Sport Mineral SPF 50, each listing zinc oxide 24.08% as the sole active with no organic filters.^[14] If benzene were a byproduct of chemical UV filters, these products could not have been affected. Mineral formulations raise their own questions, which we handled in our comparison of nano and non-nano zinc oxide, but benzene is not one of them.

Sunscreen was a small share of the affected products. The FDA enforcement database returns 114 benzene-related recall records, of which 8 are sunscreens.^[14] The rest are deodorants, hand sanitizers, acne products, and other consumer goods, with 19 records from one hand sanitizer manufacturer, 19 from Bayer, and 19 from Procter & Gamble. What links them is not sun protection.

Two limits on that account. The 2 ppm figure is a conditional limit rather than a safety endorsement, and 14 batches exceeded it. And the recalls covered specific lots rather than product lines: the Johnson & Johnson record is still listed as “Ongoing,” while the Beiersdorf and Edgewell records are marked “Terminated.”^[14]

Does sunscreen cause cancer? Where the evidence lands

The evidence here does not support a blanket safety declaration, and this article is not making one. What it supports is a comparison of two risks measured very differently.

On one side: benzene, IARC Group 1, found as a contaminant in a minority of batches concentrated in aerosol sprays, recalled by manufacturers, absent from every ingredient list. On the other: ultraviolet radiation, which the AAD notes has been declared “a known carcinogen” by the US Department of Health and Human Services and by IARC.^[15] Both are Group 1. One was recalled. The other arrives every clear morning, and sunscreen is what blocks it. The benefit side of that comparison was measured in a randomized trial, with the reductions listed above and an explicit finding of no harm; the harm side, in the same trial, was not detected. The AAD’s summary position: “Scientific evidence supports the benefits of sun protection, including using sunscreen to minimize short-term and long-term damage to the skin from the sun’s rays.”^[15]

None of this settles an individual decision. If a specific ingredient concerns you, mineral filters are an option, though the recall record shows they were not spared contamination either. If you are pregnant, we gathered what the evidence does and does not cover in our piece on pregnancy-safe sunscreen. If the practical obstacle is reapplication rather than safety, our guide to using a sun stick over makeup covers the technique. And if you have a personal or family history that makes any of this more than an abstract question, that is a conversation for your dermatologist rather than a blog post.

FAQ

Does sunscreen cause melanoma?

The only randomized trial found fewer melanomas in the daily sunscreen group, not more: 3 invasive melanomas against 11, hazard ratio 0.27 (0.08-0.97).^[3] Pooled cohort studies do report a higher melanoma rate among sunscreen users (HR 1.27, 1.07-1.51), and that number is real, but the meta-analysis reporting it attributes the pattern to “innate confounding by indication” and notes that “The only RCT showed a protective effect of sunscreen.”^[9] People who burn easily buy more sunscreen and also develop more melanoma, which pulls observational comparisons opposite to the causal direction.

Is benzene still in sunscreen?

The contaminated batches identified in 2021 and 2022 were recalled as Class I actions, covering roughly 2.6 million cans across three companies.^[14] Benzene was never a sunscreen ingredient: the FDA recorded the recalls as “Chemical Contamination,” and the same contamination turned up in deodorants, hand sanitizers, and other consumer products.^[14] Mineral formulas with zinc oxide as the only active were recalled alongside chemical ones, and non-aerosol formats were not implicated. In the FDA database the Johnson & Johnson record is still listed as “Ongoing” while the other two read “Terminated,” so this covers specific lots rather than a closed question.

Should I stop wearing sunscreen because of these concerns?

That decision belongs with your dermatologist, who knows your skin history. What the published record shows: a 1,621-person randomized trial found reduced squamous cell carcinoma, reduced invasive melanoma, fewer solar keratoses, and less photoaging, alongside an explicit statement of “no harmful effect of daily use of sunscreen in this medium-term study.”^[1][3][4][5] It found no reduction in basal cell carcinoma, so the protection is not universal. And UV radiation carries the same IARC Group 1 classification as the contaminant that prompted the recalls.^[15]

Sources:
[1] Green A, Williams G, Neale R, et al. Daily sunscreen application and betacarotene supplementation in prevention of basal-cell and squamous-cell carcinomas of the skin: a randomised controlled trial. Lancet. 1999;354(9180):723-9. PMID 10475183. Source for the SCC tumor rates (1,115 vs 1,832 per 100,000; RR 0.61, 0.46-0.81), the absence of a BCC effect, and the quotation “There was no harmful effect of daily use of sunscreen in this medium-term study.” https://pubmed.ncbi.nlm.nih.gov/10475183/
[2] van der Pols JC, Williams GM, Pandeya N, Logan V, Green AC. Prolonged prevention of squamous cell carcinoma of the skin by regular sunscreen use. Cancer Epidemiol Biomarkers Prev. 2006;15(12):2546-8. PMID 17132769. Eight years of follow-up after the trial ended; SCC tumor rate “significantly decreased by almost 40% during the entire follow-up period (rate ratio, 0.62; 95% CI, 0.38-0.99)”; BCC “tended to decrease but not significantly.” https://pubmed.ncbi.nlm.nih.gov/17132769/
[3] Green AC, Williams GM, Logan V, Strutton GM. Reduced melanoma after regular sunscreen use: randomized trial follow-up. J Clin Oncol. 2011;29(3):257-63. PMID 21135266. “Ten years after trial cessation”: 11 vs 22 new melanomas, HR 0.50 (0.24-1.02), P = .051; invasive melanoma 3 vs 11, HR 0.27 (0.08-0.97); preinvasive HR 0.73 (0.29-1.81). https://pubmed.ncbi.nlm.nih.gov/21135266/
[4] Darlington S, Williams G, Neale R, Frost C, Green A. A randomized controlled trial to assess sunscreen application and beta carotene supplementation in the prevention of solar keratoses. Arch Dermatol. 2003;139(4):451-5. PMID 12707092. Solar keratoses 24% lower in the daily sunscreen group. https://pubmed.ncbi.nlm.nih.gov/12707092/
[5] Hughes MC, Williams GM, Baker P, Green AC. Sunscreen and prevention of skin aging: a randomized trial. Ann Intern Med. 2013;158(11):781-90. PMID 23732711. Photoaging 24% less in the daily sunscreen group (relative odds 0.76, 0.59-0.98), n=903 participants under 55. https://pubmed.ncbi.nlm.nih.gov/23732711/
[6] Neale R, Williams G, Green A. Application patterns among participants randomized to daily sunscreen use in a skin cancer prevention trial. Arch Dermatol. 2002;138(10):1319-25. PMID 12374537. Median application density 0.79 mg/cm², described as “less than half the amount needed to achieve the labeled sun protection factor”; 56% applied on five or more days per week. https://pubmed.ncbi.nlm.nih.gov/12374537/
[7] Young AR, Greenaway J, Harrison GI, et al. Sub-optimal Application of a High SPF Sunscreen Prevents Epidermal DNA Damage in vivo. Acta Derm Venereol. 2018;98(9):880-887. PMID 29944164. Cyclobutane pyrimidine dimer formation measured at 0.75, 1.3, and 2.0 mg/cm²; significant reductions at 1.3 and 2.0 (p≤0.001-0.05), with the 0.75 “typical use” condition reaching significance under repeated five-day exposure by HPLC-MS/MS (p<0.001). https://pubmed.ncbi.nlm.nih.gov/29944164/
[8] Ghiasvand R, Weiderpass E, Green AC, Lund E, Veierød MB. Sunscreen Use and Subsequent Melanoma Risk: A Population-Based Cohort Study. J Clin Oncol. 2016;34(33):3976-3983. PMID 27621396. Norwegian Women and Cancer Study; 143,844 women aged 40-75, 1,532,247 person-years, 722 melanomas; SPF ≥15 vs SPF <15 HR 0.67 (0.53-0.83); the 18% figure is a population attributable fraction estimate, not an observed reduction. Also the source for “Sunscreen users reported significantly more sunburns and sunbathing vacations and were more likely to use indoor tanning devices.” https://pubmed.ncbi.nlm.nih.gov/27621396/
[9] Rueegg CS, Stenehjem JS, Egger M, et al. Challenges in assessing the sunscreen-melanoma association. Int J Cancer. 2019;144(11):2651-2668. PMID 30447006, PMC6451658. Meta-analysis of 28 studies and 21,069 melanoma cases (1979-2018). Source for all five design-specific pooled estimates in the table and for the conclusion quoted in the text. https://pubmed.ncbi.nlm.nih.gov/30447006/
[10] Iannacone MR, Youlden DR, Baade PD, Aitken JF, Green AC. Melanoma incidence trends and survival in adolescents and young adults in Queensland, Australia. Int J Cancer. 2015;136(3):603-9. PMID 24806428, PMC4277328. Joinpoint regression of Queensland registry data, 1982-2010, ages 15-24. Source for all annual percent change figures and for the quoted summary. Five-year relative survival for 2006-2010 was 95.7% (92.9-97.5). https://pubmed.ncbi.nlm.nih.gov/24806428/
[11] Whiteman DC, Neale RE, Baade P, Olsen CM, Pandeya N. Changes in the incidence of melanoma in Australia, 2006-2021, by age group and ancestry: a modelling study. Med J Aust. 2024;221(5):251-257. PMID 39217597. Source for the high-risk ancestry share falling from 85.3% to 71.1% and for the conclusion quoted in the text. The paper’s own note that annual percent changes below age 34 were “generally not statistically significant” is why we treat reference [10] as the primary evidence for the decline in young Australians. https://pubmed.ncbi.nlm.nih.gov/39217597/
[12] Lapides R, Saravi B, Mueller A, et al. Possible Explanations for Rising Melanoma Rates Despite Increased Sunscreen Use over the Past Several Decades. Cancers (Basel). 2023;15(24):5868. PMID 38136411, PMC10741796. Source for the five explanations listed, which the authors present as hypotheses rather than established causes. https://pubmed.ncbi.nlm.nih.gov/38136411/
[13] Valisure LLC. Citizen Petition on Benzene in Sunscreen and After-sun Care Products. Docket FDA-2021-P-0497-0001, filed 2021-05-24. We read the 19-page petition PDF directly. Source for the 294 batches / 69 brands testing scope, the quoted summary of 78 batches with detectable benzene, the 6.26 ppm maximum, and the 2 ppm FDA conditional limit. https://www.regulations.gov/docket/FDA-2021-P-0497
[14] US Food and Drug Administration, openFDA drug/enforcement database, queried 2026-08-14 on reason_for_recall:benzene. Source for the three Class I recall records (Johnson & Johnson Consumer, 2021-07-14, 1,437,804 cans, “Chemical Contamination: presence of benzene”; Beiersdorf, 2021-09-30, 142,236 cans; Edgewell, 2022-07-29, 139,680 aluminum spray cans), their recorded statuses, the composition of the recalled Coppertone products, and the distribution of the 114 benzene recall records across product categories. https://open.fda.gov/apis/drug/enforcement/
[15] American Academy of Dermatology Association, sunscreen and sun protection guidance. Source for the three quoted statements: on the benefits of sun protection, on absorption not implying harm, and on the classification of UV radiation as a known carcinogen by the US Department of Health & Human Services and IARC. Captured verbatim from AAD published guidance during our source review on 2026-08-14. https://www.aad.org/public/everyday-care/sun-protection/sunscreen-patients/sunscreen-faqs and https://www.aad.org/public/everyday-care/sun-protection/sunscreen-patients/is-sunscreen-safe

Topics: Clinical Evidence · Regulation & Labels · Sun Damage & Aging · Travel · All topics