Can You Trust Korean SPF Labels? What the Purito Tests Showed
The 50+ printed on your sunscreen looks like a fact about what’s in the tube. It isn’t. It’s a measurement, taken once, from a product smeared on the backs of ten volunteers in one lab on one day. And in 2020 the whole skincare internet found out what happens when someone pays to take that measurement again.
The short version: An SPF number is a lab result, not a property of the formula, and the same sunscreen can measure differently in different labs. In 2020, two independent labs measured one Korean sunscreen at roughly SPF 19 while its box said 50+.[1][2] But if you want to know whether you can trust Korean SPF labels, the honest answer isn’t about Korea at all: consumer groups have found the same gap in American, Australian, European and Japanese brands.[2] What did change since then is the test itself. In December 2024, ISO published two new ways to measure SPF that don’t need volunteers to get burned.[5]
An SPF number is a test result, not an ingredient
SPF is measured, not calculated. Testers spread 2 milligrams of sunscreen per square centimeter onto the backs of human volunteers, shine a UV lamp on them, and compare how much UV the skin can take with and without the product.[2]
Two words do a lot of work in this story, so here they are up front:
- In vivo means on living skin, with human volunteers. This is the test the label number comes from.
- In vitro means on a plate inside a machine, with no people involved.
The international standard used for labeling is called the ISO 24444 test, and the FDA’s version is very similar. Both still put sunscreen on actual human beings, in the United States, South Korea and Australia alike.[2] It costs roughly $5,000 to $10,000 to run.[2]
What the Purito tests actually measured, one by one
In late 2020, Judit Racz, who runs the ingredient database INCIDecoder, commissioned independent lab tests on Purito’s Centella Green Level Unscented Sun. The box said SPF 50+. Several numbers came out of that testing, and they are not interchangeable.[1]
| Number | Where it came from | Type of test |
|---|---|---|
| SPF 50+, marketed as SPF 84.5 | The label and the brand’s own claim | Not disclosed in the sources we read |
| SPF 10.3 | The BASF calculator, a rough estimation tool that Racz says formulators know to be on the conservative side | Computer model, not a lab test |
| SPF 15.8, standard deviation 2.3 (UVA-PF 9.1 ± 1) | Ita-Test, a Polish lab | In vitro |
| SPF 19 (mean) | Polish lab, 10 people, ISO 24444 | In vivo |
| SPF 19.2 ± 2.4 | A German lab, 5-person screening version | In vivo |
| SPF 28.4 | Korea Institute of Dermatological Sciences, a separate independent test | Method not stated in our source |
Why did Racz test it in the first place? Purito had disclosed unusually low filter amounts, 3% Uvinul A and 2% Uvinul T, and said the SPF 50+ was SPF 84.5.[1] Racz was careful about what that did and didn’t prove: “Though I do not say it is theoretically impossible to create a broad-spectrum high SPF product with only two organic filters, most brands use a wider array of filters and at a 15% to 20% percentage to create an SPF 50+ product.”[3] Note the hedge. A short filter list isn’t evidence of anything on its own.
Purito halted sales of three of its sunscreens and ordered fresh in vitro and in vivo testing done outside its manufacturer.[1][3][4] The brand also described how the product got made in the first place: “When developing a product as an ODM (original design manufacturing) system several parties are involved such as the manufacturing company, the brand, KFDA, and the testing lab. As a brand we have requested the manufacturer to develop an exclusive product for PURITO, for which we received the formulation.”[4]
One thing worth pinning down before we go on: the product in that table is the Centella Green Level Unscented Sun, a formula that got pulled. It is not the sunscreen Purito sells today, and we kept those two separate when we looked at the brand’s current formula.
How far can one sunscreen move between labs?
Further than most people would guess. Procter & Gamble sent the same SPF 100 sunscreen to five different labs, told them the SPF was somewhere between 20 and 100, and got back five results ranging from 37 to 75.[2] Same product, same standard, roughly double the answer depending on who ran it.
Chemistry PhD Michelle Wong lists several reasons the number moves:
- Labs are influenced by what you tell them. When labs were given an expected SPF, they were more likely to return it. Wong pointed out that the lab in the INCIDecoder test had been told to expect around 20.[2]
- The method had loose ends. ISO 24444 was revised in 2019 to nail down things that hadn’t been specified, including at least 15 drops per 30 square centimeters, a spreading time of 30 to 40 seconds, and color vision checks for the people grading whether skin has burned.[2]
- Batches drift. A brand usually runs the label test on an early version of a formula and might not repeat it for years.[2]
So why did the Purito result count as more than noise? Because of a threshold Wong describes: “I think when you have two in vivo SPF tests that agree and one that doesn’t, that’s when you can confidently say there probably is a problem somewhere, and the burden of proof switches.”[2] Two independent in vivo tests, run in two countries, landed at 19 and 19.2 against a label of 50+ and a published figure of 84.5.
Wong was equally clear about the limits of the tests that raised the alarm. Neither one was consistent enough to be used for a label, and one of them ran on only five people, where a label result has to hold up across at least 10. Wong still judged them good enough to raise doubt, because they were in vivo and used largely the same procedures.[2] And Wong didn’t call it fraud: “I don’t think there’s necessarily any wrongdoing on anyone’s part here.”[2]
Am I supposed to distrust Korean sunscreen now?
No, and the reason is that the same thing keeps turning up everywhere anyone bothers to retest. Consumer organizations that buy sunscreens off the shelf and measure them again have published these results:
| Who tested | When | What they found |
|---|---|---|
| Hong Kong Consumer Council | October 2020 | 25 of 30 sunscreens came in under their SPF label. Brands that missed included Estée Lauder, Shiseido, Anessa, Curél and Laneige[2] |
| Consumer Reports (US) | 2016 | 23 of 60 tested at less than half their labeled SPF[2] |
| Choice (Australia) | Late 2015 | Of six SPF 50+ sunscreens, two met their claim[2] |
| European consumer tests | Reported 2020 | 15% of sunscreens tested fell short of their label claim[2] |
The lowest result in the Hong Kong round wasn’t Korean. It was Fancl, a Japanese brand, labeled SPF 50+ and measured at SPF 14.3.[2] In the American round, two SPF 50 kids’ lotions, one from Banana Boat and one from CVS, came out at SPF 8.[3] And the biggest sunscreen labeling scandal on record involves AMA Labs, one of the biggest US sunscreen testing labs, whose indictment alleges fraud over 30 years, from 1987 to 2017.[2]
If there’s a pattern here, industry people in 2020 argued it was about formulas, not passports. Maria Semykoz, co-founder of What’s In My Jar, put it this way: “The less UV filters a formula uses, the more cosmetically elegant the sunscreen will feel, but, unfortunately, it also makes it less likely to provide a high level of sun protection.”[3] That’s her reading of a tendency, and it’s worth holding loosely, because Wong’s own point is that you can’t confidently judge a label from filter percentages alone.[2] Light textures are a design goal in a lot of markets, and the same brand often ships a different filter set depending on where it’s sold.
What actually changed after 2020
The test did. In December 2024, ISO published two new methods for measuring SPF, and both are available to industry “as alternatives to the current ISO 24444 method.”[5]
- ISO 23675, the Double Plate Method, is fully in vitro. It measures SPF with no human volunteers at all, and it was led by Cosmetics Europe and scientists from its member companies.[5]
- ISO 23698 uses Hybrid Diffuse Reflectance Spectroscopy, an optical reading taken on skin that characterizes SPF, UVA protection and critical wavelength without needing a biological response.[5]
Both were developed with extensive work to make sure they correlate with in vivo results from ISO 24444.[5] This work had been running for years: back in 2020, Wong was already describing the in vitro plate method as being rolled out but not yet carrying an official tick of approval.[2] Four years later it got one.
Two honest limits, because they’re in the same announcement. The Double Plate Method applies to emulsions and alcoholic one-phase products, and not to loose or pressed powders or sticks. It also can’t be used to determine water resistance, which is a separate test.[5] And the trade body’s own framing is that no single measurement decides an SPF: “companies do not rely just on one measurement for the SPF value: a weight of evidence is produced.”[5]
Can I trust Korean SPF labels when I’m standing in a shop?
Treat the number as a good estimate rather than a guarantee, then lean on the things you can see with your own eyes. Racz, who paid for the tests that started all this, was blunt about the ceiling on certainty: “There is no way for the consumers to know for sure that a sunscreen meets the SPF claims on its label, but, in general, when it comes to organic filters, if a wider range of filters are used and in higher concentrations, there is a better chance of high SPF protection.”[3]
Wong’s practical list points the same direction. Things that tilt the odds toward a sunscreen doing what it says:[2]
- A stated water resistance rating.
- A higher total concentration of UV filters.
- A thicker product, since sticks tend to outperform lotions when you apply enough.
- A brand big enough to attract attention from regulators.
- Good storage, bought from a retailer you’d trust with anything else.
And then there’s the factor that dwarfs all of it, which is how much you put on. A thin layer cuts your real protection roughly in half no matter whose label you’re reading, so the difference between a good number and a great one matters much less than using enough of it. We ran that arithmetic on SPF 30 against SPF 50 and the answer was the same. Wong’s closing note is worth keeping too: sunscreen is one part of sun protection, alongside a hat, shade, and staying out of the midday sun.[2]
FAQ
Do I need to stop using my Korean sunscreen because of the Purito story?
No. The 2020 tests covered a single formula, Purito’s Centella Green Level Unscented Sun, which the brand pulled from sale. Nothing in those results describes any other sunscreen. And the same shortfall has shown up in American, Australian, European and Japanese products when consumer groups retested them, so this was never a story about one country.
Why do I keep seeing different SPF numbers for the same sunscreen?
Because SPF is a measurement and measurements vary. Procter & Gamble sent one SPF 100 sunscreen to five labs and got results from 37 to 75. Batch differences, the mix of volunteers, and even telling a lab what to expect can all move the number. That’s why a single low result usually isn’t treated as proof of anything.
Am I safer buying a Western sunscreen instead?
Not automatically. Consumer Reports found 23 of 60 sunscreens tested at under half their labeled SPF in 2016, including two SPF 50 kids’ lotions that measured SPF 8. Australia’s Choice tested six SPF 50+ products in 2015 and two met their claim. The biggest testing fraud case on record involved a US lab.
What should I actually look for if I can’t verify the SPF myself?
Look for a water resistance rating, a higher total concentration of UV filters, a thicker texture, a brand large enough that regulators notice it, and a retailer you trust. Racz adds that a wider range of filters at higher concentrations gives a better chance of high protection, though she’s clear there’s no way to be certain.
Am I going to see more reliable SPF numbers in future?
The testing is moving that way. In December 2024, ISO published ISO 23675, a fully in vitro method needing no volunteers, and ISO 23698, an optical method read on skin. Both are offered as alternatives to the human-volunteer ISO 24444 test, and both were built to correlate with it.
Sources
The 2020 tests and the brand response (read in full 2026-07-28):
[1] Racz J, “Purito Centella Unscented Sun tests as SPF 19 in two different European labs,” INCIDecoder, 3 December 2020, updated 22 December 2020. Discloses 3% Uvinul A and 2% Uvinul T and the brand’s SPF 84.5 claim; BASF calculator estimate SPF 10.3; Ita-Test (Poland) in vitro SPF 15.8, standard deviation 2.3, UVA-PF 9.1 ± 1; 10-person in vivo test under ISO 24444, mean SPF 19; German lab 5-person screening in vivo SPF 19.2 ± 2.4; update notes suspension of three Purito sunscreens and a Korea Institute of Dermatological Sciences test measuring SPF 28.4. link
[4] Bargh B, “Ultra-denial it: Purito shifts blame for SPF scandal onto manufacturers,” Cosmetics Business, 7 December 2020. Quotes Purito’s statement on the ODM development process and reports that Purito requested in vitro and in vivo tests for all three sunscreen products. link
[3] Brown R, “After Independent Testing Discovers Purito’s Mistaken SPF Claim, How Can Sun Care Brands Regain Consumers’ Trust?”, Beauty Independent, 10 December 2020. Source of the full Racz quotes, the Semykoz quote on cosmetic elegance versus filter load, and the Consumer Reports finding that Banana Boat and CVS SPF 50 kids’ lotions measured SPF 8. link
How SPF testing works and how much it varies (read in full 2026-07-28):
[2] Wong M, “Purito Sunscreen and All About SPF Testing,” Lab Muffin Beauty Science, 10 December 2020. Source of the 2 mg per square centimeter dose, the $5,000 to $10,000 test cost, the ISO 24444 and FDA in vivo requirement across the US, South Korea and Australia, the Procter & Gamble five-lab range of 37 to 75, the 2019 ISO 24444 revision details, the 10-volunteer consistency requirement, the two-agreeing-tests threshold, the Hong Kong (25 of 30), Consumer Reports (23 of 60), Choice Australia (2 of 6) and European (15%) retest figures, Fancl at SPF 14.3, the AMA Labs indictment covering 1987 to 2017, and the shopper checklist. link
The new SPF test methods (read in full 2026-07-28):
[5] “ISO Publishes New SPF Test Methods,” CTPA, 14 January 2025. Reports that ISO published ISO 23675 (Double Plate Method, in vitro, led by Cosmetics Europe) and ISO 23698 (Hybrid Diffuse Reflectance Spectroscopy) in December 2024 as alternatives to ISO 24444, that both were developed to correlate with ISO 24444 in vivo assessments, that the Double Plate Method excludes powders and sticks and cannot determine water resistance, and that companies produce a weight of evidence rather than relying on one measurement. link
