In this article
The short answer
In the UK and EU a sunscreen is a cosmetic, and the SPF on the front is a claim you must be able to prove with a test on the finished product. The benchmark is Commission Recommendation 2006/647/EC: SPF measured by a recognised ISO method (in vivo ISO 24444, or since December 2024 the in vitro Double Plate Method or HDRS), UVA protection at least a third of the labelled SPF with a critical wavelength of 370 nm, the SPF shown in a fixed series with its category, and water resistance claimed only after an immersion test. Sunblock, total protection and all day claims are out, application and reapplication directions are in, and only Annex VI filters may be used, with GB and EU limits now different for several of them.
Key takeaways
- You cannot work out an SPF from the filter percentage. Two creams with the same zinc oxide level can test at SPF 8 and SPF 25.
- ISO 24444 on volunteers is still the reference test, but the in vitro Double Plate Method (ISO 23675) and HDRS (ISO 23698) have been ISO standards since December 2024 and Cosmetics Europe now recommends phasing them in.
- Every sunscreen should reach a UVA protection factor of at least one third of its labelled SPF and a critical wavelength of 370 nm. The circled UVA logo is how you show it.
- Labels use a fixed series: 6 and 10 (low), 15, 20 and 25 (medium), 30 and 50 (high) and 50+ (very high, measured 60 or more). Round down, never up.
- Water resistant means at least half the SPF survives 2 x 20 minutes of immersion. Very water resistant means 4 x 20 minutes. Waterproof is not a claim you can support.
- GB and EU filter lists have split: 4-MBC is now banned in both, but homosalate is face-only at 7.34% in the EU and still 10% in GB.
- Zinc oxide and nano titanium dioxide must not be used where they can be inhaled, so check every spray and loose powder against the Annex VI conditions.
The moisturiser had 15 percent zinc oxide in it. The maker had read somewhere that 15 percent zinc gives roughly SPF 15, so SPF 15 went on the front of the jar. When the brand finally sent it to a lab, it measured SPF 9. Nobody had lied on purpose, but the label was still wrong, and with sun care a wrong label means someone gets burned.
That is why SPF is the most evidence-hungry claim in cosmetics. This guide covers what has to sit behind an SPF number in the UK and EU in 2026: how SPF and UVA are measured now that two new ISO methods exist, which numbers and words a label may carry, what water resistant really requires, where Great Britain and the EU now disagree on filters, and what your safety assessor needs to see. If your question is mainly about which filters are allowed, our UV filters guide covers Annex VI in more depth.
The legal footing for an SPF claim
Sunscreens are cosmetics on both sides of the Channel. In the EU and Northern Ireland the framework is Regulation (EC) No 1223/2009. In Great Britain it is the assimilated GB version of the same Regulation, enforced through the Cosmetic Products Enforcement Regulations 2013. Article 20 requires claims to be truthful and supported, and Regulation (EU) No 655/2013 sets the six common criteria every claim is judged against.
The detail for sun products comes from Commission Recommendation 2006/647/EC. A recommendation is not binding law, but market surveillance authorities in the EU use it as the yardstick for whether an SPF claim is honest, and the UK industry body CTPA advises that it applies in the UK too. In practice, if your label departs from it, you will need a very good reason.
SPF and UVA measure different things
SPF is about UVB, the shorter wavelengths that cause sunburn. Formally it is the ratio of the UV dose that produces minimal redness on protected skin to the dose that does so on unprotected skin. UVA is the longer band, 320 to 400 nm. It penetrates deeper, drives photoageing and contributes to skin cancer, but does not redden the skin quickly, so an SPF test barely sees it.
That is the whole reason the rules ask for two sets of numbers. A product can score SPF 50 and still give weak UVA cover. So a sunscreen needs a measured SPF, a measured UVA protection factor and a measured critical wavelength, the point below which 90 percent of the product's absorbance between 290 and 400 nm sits.
How SPF is tested in 2026
For two decades the answer was simple: ISO 24444, on human volunteers, with the product applied at 2 mg/cm² and at least ten valid subjects. That changed in December 2024, when ISO published two alternatives. The Recommendation has always said in vitro methods should be preferred where they give equivalent results, and Cosmetics Europe Recommendation No 26 (revised June 2025) now tells its members to start phasing the new methods in. Products already tested in vivo are not expected to be retested.
| Method | What it measures | Where it fits |
|---|---|---|
| ISO 24444:2019 (in vivo) | SPF on volunteers' skin | Still the reference method; any format; needed alongside ISO 16217 for water resistance |
| ISO 23675:2024 Double Plate Method (in vitro) | SPF on two types of plate, no volunteers | Emulsions and one-phase alcoholic products; not loose or pressed powders, sticks or water resistance |
| ISO 23698:2024 HDRS (hybrid) | SPF, UVA protection factor and critical wavelength, optically on skin without inducing a reaction | Emulsions and single-phase products; not yet evaluated for powders |
| ISO 24443 (in vitro) | UVA protection factor and critical wavelength | The usual UVA test in Europe |
| ISO 24442 (in vivo) | UVA protection factor by persistent pigment darkening | Alternative UVA route |
Whichever route you choose, CTPA's advice is to pick the method that suits the formula and build a documented dossier behind the final numbers. Ask the lab what experience it has with the newer methods. They are only as good as the training of the people running them.
Why the filter percentage tells you nothing
SPF depends on the filter, but also on its particle size and coating, how evenly it is dispersed, the emollients and film formers around it, the pH and the way the product spreads and dries. Two creams with identical zinc oxide levels can test at SPF 8 and SPF 25. Supplier calculators that predict SPF from filter levels are useful for formulating and worthless as evidence. Your assessor and a Trading Standards officer will treat them the same way.
In practice
Budget the SPF test as part of developing the product, not as an extra at the end. Until it is tested, what you have is a moisturiser with a number printed on it.
Labelled SPF numbers and protection categories
The Recommendation limits the numbers you can print and asks for the category to be shown at least as prominently as the SPF itself. You label down to the nearest step in the series. A measured 24 is labelled SPF 20. A measured 58 is SPF 50. Only a measured 60 or more earns 50+.
| Labelled category | Labelled SPF | Measured SPF | Minimum UVA |
|---|---|---|---|
| Low protection | 6 | 6 to 9.9 | UVA protection factor at least one third of the labelled SPF, critical wavelength at least 370 nm |
| Low protection | 10 | 10 to 14.9 | Same |
| Medium protection | 15 / 20 / 25 | 15 to 19.9 / 20 to 24.9 / 25 to 29.9 | Same |
| High protection | 30 / 50 | 30 to 49.9 / 50 to 59.9 | Same |
| Very high protection | 50+ | 60 or more | Same |
Below SPF 6 a product should not claim sun protection at all. And there is no SPF 70 or SPF 100 on a UK or EU label, however high the test came out.
The UVA minimum and the UVA logo
Many brand owners think the one-third rule is just the price of the logo. It is not. The Recommendation says every sunscreen should protect against both UVB and UVA, with a UVA protection factor of at least one third of the labelled SPF and a critical wavelength of at least 370 nm. An SPF 30 therefore needs a UVA protection factor of 10 or more.
The circled UVA symbol is a standard label issued by Cosmetics Europe to show that minimum is met. It is not legally required, but a sunscreen without it invites the obvious question, and retailers now expect it. You cannot use it because zinc oxide is known to absorb UVA. It has to be measured on your formula. Some UK retailers also run their own UVA star schemes, which are separate and licensed.
Water resistant and very water resistant
These are tested claims with fixed meanings. Under Cosmetics Europe Recommendation No 16 the SPF is measured before and after immersion using ISO 16217, and retention is calculated with ISO 18861. Water resistant means at least 50 percent of the SPF is retained after two 20-minute immersions, 40 minutes in total. Very water resistant means the same after four immersions, 80 minutes in total. The figure is based on the lower confidence limit, not a lucky average.
The Double Plate Method cannot do this, so water resistance still means volunteers. Waterproof and sweatproof are not claims anyone can substantiate, because no product stays fully effective in water. Our guide to waterproof cosmetic claims explains how regulators read those words more generally.
Claims you cannot make
- "Sunblock", "sunblocker", "total protection", "100% protection". The Recommendation rules out anything implying complete protection.
- "All day protection", "no need to reapply". Also excluded. Every sunscreen needs reapplying.
- "Waterproof", "sweatproof". Use the tested water resistant wording instead.
- "Prevents skin cancer", "repairs sun damage", "prevents actinic keratosis". These are medical claims and pull the product out of cosmetics territory.
- "Reef safe", "ocean friendly". Not banned as such, but there is no agreed test behind them. In the EU, Directive (EU) 2024/825 has tightened the rules on environmental claims from 27 September 2026, and the CMA polices green claims in the UK. Treat them as claims that need evidence like any other.
What every sunscreen label should say
The Recommendation asks for three kinds of statement, and enforcement officers check for all of them:
- Warnings that it is not 100 percent protection, with advice such as "Do not stay too long in the sun, even while using a sunscreen product", "Keep babies and young children out of direct sunlight" and "Over-exposure to the sun is a serious health threat".
- Directions for use, such as "Apply before sun exposure" and "Reapply frequently to maintain protection, especially after perspiring, swimming or towelling".
- How much to apply, by pictogram, illustration or measure, with a warning like "Reducing this quantity will lower the level of protection significantly".
Add any warning the Annex VI entry requires for your filters (benzophenone-3 used as a UV filter, for instance, needs "Contains Benzophenone-3" in the EU) and the (nano) suffix after any nanomaterial in the ingredient list. A jar with SPF 50 in large type and no application guidance is non-compliant however good its test report is. Our labelling guide covers the general label rules that sit alongside these.
UV filters: Annex VI, and where GB and the EU now differ
Only filters on Annex VI may be used as UV filters, at their listed maximums and conditions. A filter approved in the US, Australia or Japan is not usable here unless it is on that list. Since 2021 the EU and GB lists have been amended separately, and for sun care the gap now matters:
| Filter | EU and Northern Ireland | Great Britain |
|---|---|---|
| 4-Methylbenzylidene Camphor (4-MBC) | Prohibited by Regulation (EU) 2024/996: no new products from 1 May 2025, off the market from 1 May 2026 | Prohibited by SI 2026/23 from 15 July 2026; stock placed before then may be sold until 14 January 2027 |
| Homosalate | Face products only, up to 7.34%, not in propellant sprays (full application from 1 July 2025) | Still up to 10% in all products (CTPA) |
| Benzophenone-3 | 6% face, hand and lip; 2.2% body including sprays; 0.5% other (Regulation (EU) 2022/1176) | Same tiers since 21 January 2026 (SI 2025/901); older compliant stock could be sold until 21 July 2026 |
| Octocrylene | 9% in propellant sprays, 10% in other products | 10% |
The practical upshot is that a body lotion with 10 percent homosalate can be legal in Manchester and not in Dublin or Belfast. If you sell in both markets, the assessor checks both lists, and the simplest answer is usually one formula built to the stricter limit.
Sprays, powders and nano filters
Format can rule a filter out. The Annex VI entries for zinc oxide, in both its standard and nano forms, and for nano titanium dioxide say they must not be used in applications that may expose the user's lungs by inhalation, which in practice means sprays and loose powders. Several organic filters carry the same condition, and tris-biphenyl triazine is not to be used in sprays at all. Read the Annex VI conditions for each filter against your actual pack, not just the formula.
Nano filters also have to meet the particle specifications written into their entries, so ask your supplier for a statement that the grade complies. Because they are listed in Annex VI, they do not need the separate six-month nanomaterial notification under Article 16, but they still need the (nano) suffix on the label.
SPF moisturisers, foundations and lip balms
The 2006 Recommendation defines a sunscreen as a product used exclusively or mainly to protect against UV. A day cream, tinted moisturiser or foundation with SPF is what the industry calls a secondary sun product. Strictly, the Recommendation was not written for these, but that is not a loophole. The moment the pack shows an SPF number, you have made a sun protection claim, it must be substantiated under Article 20, the filters must come from Annex VI, and assessors and enforcement officers expect the same test methods and number series.
If you are not ready to test, sell it as a moisturiser and leave the number off. Mentioning that it contains zinc oxide is not a protection claim in itself. A lot of brands launch that way and add a tested SPF version later, which is a sensible order. An SPF lip balm is also a lip product, so the face, hand and lip limits apply to filters such as benzophenone-3 and the exposure assessment allows for some ingestion. Our lip balm CPSR guide covers that side.
How the CPSR handles a sunscreen
A sunscreen needs a Cosmetic Product Safety Report like any other cosmetic, but three things stand out. First, the assessor checks every filter against Annex VI for the market concerned: concentration, form, nano specification and conditions. Second, exposure is high. The SCCS Notes of Guidance use 18.0 g a day of sunscreen lotion over the whole body surface of about 17,500 cm² for margin of safety calculations, far above a face cream, so filters and preservatives are tested against that load.
Third, the claims. The assessor does not run the SPF test and cannot infer an SPF from the formula. What they do is confirm that the SPF, UVA and any water resistance claims are backed by reports on the finished product, and those reports go into the Product Information File as the proof of effect required by Article 11. Stability matters more than usual too, because filters can drift out of suspension or degrade. See our note on stability testing.
The 2006 Recommendation is being rewritten
The Commission started revising the Recommendation in June 2024, and a working group of member state authorities, the Commission and industry has been through several drafts. At a meeting on 22 September 2026 there was broad agreement that secondary sun products such as SPF day creams and make-up should be identifiable to consumers, probably through a new symbol that will be consumer-tested. Updated efficacy criteria linked to the new test methods and firmer usage instructions are also on the table. France's agency ANSES has argued for tighter rules still, including scrapping the very high category.
As of November 2026 nothing has been adopted, and a transition period is expected once it is. Keep labelling to the 2006 text for now, but if you are designing packaging for 2027 and beyond, leave room on the pack. Our round-up of UK regulation changes in 2026 tracks the GB side.
A worked example: the zinc moisturiser, done properly
Back to the jar that said SPF 15 and tested at SPF 9. The brand reworks the base with a better-dispersed coated zinc oxide grade, keeps it at 15 percent (well inside the 25 percent Annex VI maximum) and runs a short stability study first, because there is no point testing a formula that separates in a month.
The lab tests it by the Double Plate Method, since it is an emulsion and no water resistance claim is planned. It measures SPF 18, which labels as medium protection, SPF 15. The ISO 24443 UVA test gives a UVA protection factor of 6.5 and a critical wavelength of 372 nm. A third of 15 is 5, so the product meets the UVA minimum and can carry the circled logo.
The zinc grade is nano and the product is in a pump bottle, not a spray, so it is permitted, and the ingredient list reads Zinc Oxide (nano). The back label gets the category, application and reapplication directions, the reduced quantity warning and the sun-safety advice. Because it is a daily face cream, the brand also adds a line that it is not a substitute for a dedicated sunscreen during prolonged sun exposure, which is where the revised Recommendation looks likely to land anyway. The CPSR checks the zinc grade against Annex VI, uses the sunscreen exposure values, and lists both lab reports as the evidence for each claim. Now the number on the front is true.
Ask yourself
If a customer who burned wearing your product asked to see the SPF report, what would you send them? If the honest answer is a supplier's estimate, the number on your label is a promise you cannot keep.
Costs and lead times
An in vivo SPF test is the most expensive routine test most small brands will ever buy, because it needs volunteers, a solar simulator and a tight protocol. The in vitro methods are typically quicker and cheaper per test, and they can run all year without worrying about tanned volunteers. UVA in vitro testing is cheaper still. Water resistance adds a further in vivo study.
For a first sun product, expect the testing to cost more than the CPSR and PIF together, and expect at least one reformulation. First formulas rarely land where the supplier's table said they would. Get quotes and lead times from two labs before you promise a launch date to a retailer.
Plan your SPF launch before the first test batch
Send us the formula, filter grades and the claims you want on the pack. We check each filter against the GB and EU Annex VI, tell you which tests and categories the claims need, review your label wording and prepare the CPSR and PIF once the reports are in.
Related reading
- UV filters in cosmetics: Annex VI and sunscreen rules
- Waterproof cosmetic claims: testing and substantiation
- Cosmetic claims: the six common criteria
- Prohibited cosmetic claims in the UK and EU
- How to create a compliant cosmetic label
- CPSR for lip balm
- Cosmetic stability testing explained
- UK cosmetic regulation changes in 2026
- How to prepare for a CPSR: the document checklist
- All regulatory compliance articles
Myths we hear
- Fifteen percent zinc oxide gives SPF 15. Only a test on the finished product gives a number. Particle size, coating, dispersion and the film the base leaves on skin move the result more than the percentage does.
- Zinc is broad spectrum, so I can use the UVA logo. The logo needs a measured UVA protection factor of at least a third of the labelled SPF and a critical wavelength of 370 nm on your formula, not the filter's reputation.
- A day cream with SPF is not really a sunscreen, so the rules do not apply. The 2006 Recommendation was written for products mainly for sun protection, but the SPF number is still a claim that needs the same quality of evidence. Regulators are now drafting specific rules for these secondary products.
- If I test it at SPF 100, I can print SPF 100. The labelled series stops at 50+. A measured 60 and a measured 110 both go on the pack as 50+, very high protection.
- The new in vitro methods mean I can skip testing. They replace the volunteers, not the test. You still need a report on your finished product from a lab competent in the method, and water resistance still needs an in vivo study.
What to do this week
- List every product in your range that shows an SPF number or a UVA logo and match each one to a test report on that exact formula and pack.
- Take the number off anything without a report, or relaunch it as a moisturiser with no sun protection claim until the test is done.
- Check every UV filter against both the GB and the EU Annex VI if you sell in both, paying particular attention to homosalate, benzophenone-3 and any 4-MBC stock.
- Confirm that no zinc oxide or nano titanium dioxide grade sits in a spray or loose powder, and get the supplier's nano specification on file.
- Ask two labs for quotes and lead times on ISO 24444 or ISO 23675, ISO 24443 and, if you want a water resistance claim, ISO 16217 with ISO 18861.
- Rewrite the back label with the category, the application and reapplication directions, the reduced-quantity warning and the sun-safety advice before you order print.
Frequently asked questions
Can I calculate SPF from the amount of zinc oxide or other filters in my formula?
No. SPF depends on the whole formula: filter grade and particle size, how well it is dispersed, the emollients and film formers and how evenly the product spreads. Supplier SPF calculators are formulation aids, not evidence. A labelled SPF must come from a test on the finished product, using ISO 24444 on volunteers or one of the two newer ISO methods.
Which SPF test methods are accepted in the UK and EU in 2026?
The law does not name a method, but Commission Recommendation 2006/647/EC expects standardised, reproducible ISO methods and prefers in vitro ones. ISO 24444:2019 (in vivo) remains the reference. ISO 23675:2024, the in vitro Double Plate Method, and ISO 23698:2024, hybrid diffuse reflectance spectroscopy, were published in December 2024 and Cosmetics Europe recommends phasing them in. The Double Plate Method does not cover powders, sticks or water resistance.
What is the UVA requirement and when can I use the UVA logo?
The Recommendation says every sunscreen should give a UVA protection factor of at least one third of its labelled SPF and a critical wavelength of at least 370 nm. The circled UVA logo, issued by Cosmetics Europe, shows the product meets that minimum. Both values must be measured on the finished product, usually by ISO 24443 in vitro or ISO 24442 in vivo, and an SPF 30 therefore needs a UVA protection factor of at least 10.
What SPF numbers can go on a sunscreen label?
Only the recommended series, shown with its category: 6 and 10 (low protection), 15, 20 and 25 (medium), 30 and 50 (high) and 50+ (very high). Each labelled number covers a measured band, so a measured SPF of 24 is labelled 20, and 50+ needs a measured SPF of at least 60. The category should be at least as prominent as the number.
Which sunscreen claims are not allowed?
Anything implying 100 percent protection, such as sunblock, sunblocker or total protection, and anything implying you never need to reapply, such as all day protection. Waterproof and sweatproof overstate what any product can do and fail the honesty criterion in Regulation (EU) No 655/2013. Claims about preventing skin cancer or treating sun damage push the product towards medicines law.
What statements should a sunscreen label carry?
Directions to apply before sun exposure and to reapply frequently, especially after sweating, swimming or towelling; guidance on how much to apply and a warning that using less lowers protection significantly; and advice that no sunscreen gives full protection, such as not staying too long in the sun and keeping babies and young children out of direct sunlight. Add any filter warning from Annex VI and the (nano) suffix where it applies.
How do I substantiate a water resistant sunscreen claim?
Test it. Under Cosmetics Europe Recommendation No 16 the SPF is measured before and after water immersion using ISO 16217 and retention is calculated by ISO 18861. Water resistant needs at least 50 percent of the SPF retained after two 20-minute immersions; very water resistant needs the same after four. The current in vitro Double Plate Method cannot be used for this.
Is an SPF moisturiser or foundation regulated like a sunscreen?
Partly. The 2006 Recommendation was written for products used mainly for sun protection, so a day cream or foundation with SPF is a secondary sun product. The SPF is still a claim, though, and must be backed by the same kind of test, filters must come from Annex VI and the number should follow the same series. The revision of the Recommendation under way in 2026 is expected to treat these products separately, possibly with their own symbol.
Official sources and legislation
- Commission Recommendation 2006/647/EC on the efficacy of sunscreen products and the claims made relating thereto
- European Commission: sunscreen products (UVA logo and standardisation)
- Regulation (EC) No 1223/2009 on cosmetic products (EUR-Lex)
- GB Cosmetics Regulation (assimilated Regulation 1223/2009, legislation.gov.uk)
- GB Annex VI: UV filters allowed in cosmetic products (legislation.gov.uk)
- The Cosmetic Products Enforcement Regulations 2013 (SI 2013/1478)
- Regulation (EU) No 655/2013: common criteria for cosmetic claims
- Regulation (EU) 2022/1176: benzophenone-3 and octocrylene limits in Annex VI
- Regulation (EU) 2024/996: 4-Methylbenzylidene Camphor prohibited
- SI 2025/901: GB benzophenone-3 restriction from 21 January 2026
- SI 2026/23: GB prohibition of 4-Methylbenzylidene Camphor from 15 July 2026
- CTPA: new homosalate restrictions do not apply in GB
- CTPA: ISO publishes new SPF test methods (ISO 23675 and ISO 23698)
- Cosmetics Europe Recommendation No 26: alternative methods to ISO 24444 (revised June 2025)
- Cosmetics Europe Recommendation No 16: water resistance labelling (revised June 2023)
- SCCS Notes of Guidance, 12th revision (SCCS/1647/22)
- Directive (EU) 2024/825: empowering consumers for the green transition (environmental claims)
- ANSES opinion supporting the revision of the sunscreen Recommendation
Join the conversation