SPF & PA+++ Testing: ISO 24444 In Vivo vs In Vitro Method & Critical Wavelength

Overview #

SPF numbers are not just marketing claims — they are test results, and the method you use to generate them determines whether your number holds up at customs, on shelf, or in a lawsuit. We test both ways in our lab: ISO 24444 in vivo on human subjects and in vitro via spectrophotometric methods. They don’t always agree. Understanding why, and when each method is appropriate, is the first thing we walk brand partners through before we finalize a sunscreen brief.

The PA+++ system adds another layer. Critical wavelength is the gating metric for broad-spectrum claims in most markets, and it’s where a lot of formulations quietly fail without anyone catching it until the third-party lab report comes back. We’ve seen finished formulas with SPF 50 on paper that couldn’t clear a 370 nm critical wavelength threshold. That’s a reformulation, not a label fix.

This guide covers how we approach SPF and PA+++ testing method selection, what drives divergence between in vivo and in vitro results, and the stability and compatibility factors that determine whether your SPF claim survives real-world conditions.

ISO 24444 In Vivo Testing: What the Method Actually Demands #

ISO 24444 is the gold standard for SPF determination. The protocol requires application at 2 mg/cm² on human panelists, UV exposure across a defined dose range, and MED (minimal erythemal dose) comparison between protected and unprotected skin. The result is a geometric mean SPF across a minimum of 10 subjects, with a coefficient of variation below 20%.

That 2 mg/cm² application rate is where most real-world SPF performance falls apart. Consumers apply somewhere between 0.5 and 1.0 mg/cm² on average. The actual in-use SPF is roughly the square root of the labeled value at half application — so your SPF 50 is performing closer to SPF 7 on a typical consumer. We tell brand partners this upfront. It doesn’t change the regulatory requirement, but it should inform how you position the product.

In our experience running in vivo panels, the biggest source of variability isn’t the UV source — it’s the panelist skin type and the technician applying the product. We’ve had batches where the same formula returned SPF 47 and SPF 53 across two separate test runs at the same CRO. Both passed SPF 50 labeling thresholds, but the spread is uncomfortable. The EU Cosmetics Regulation 1223/2009 requires SPF labeling to follow the COLIPA/ISO 24444 method for EU market claims, and the labeled value must be the lower of the 95% confidence interval — not the mean. That detail catches brands off guard.

For markets like the US, the FDA Cosmetics Guidelines treat sunscreens as OTC drug products, which means the FDA monograph method (not ISO 24444) is technically required. We flag this every time a brand wants a single global formula. You can run both methods, but they’re not interchangeable, and the results will differ.

In Vitro Methods and Critical Wavelength: The PA+++ Story #

In vitro SPF testing uses a spectrophotometer to measure UV transmittance through a thin film of product applied to a roughened PMMA plate. The method is faster, cheaper, and doesn’t require human subjects. It’s also less predictive of real skin performance because PMMA surface roughness doesn’t replicate skin texture, and film uniformity on a flat substrate is easier to achieve than on living skin.

Where in vitro really earns its place is critical wavelength and UVA ratio testing. Critical wavelength (λc) is defined as the wavelength at which the cumulative area under the absorbance curve reaches 90% of the total integral from 290 to 400 nm. For broad-spectrum claims, you need λc ≥ 370 nm. For PA+++ under the Japanese JCIA method — which is what most Asian markets and many EU brands now reference — you need a UVA-PF of at least 8, and PA++++ requires UVA-PF ≥ 16.

Here’s where formulation decisions get consequential. Zinc oxide is your best tool for pushing critical wavelength above 370 nm. At 10% uncoated ZnO, we typically see λc around 375–378 nm. Titanium dioxide alone, even at 15%, often lands at 365–368 nm — below threshold. The combination matters. In our lab, a 5% TiO₂ / 5% ZnO blend consistently clears 370 nm, while 10% TiO₂ alone frequently doesn’t. That’s not a minor formulation detail. That’s the difference between a broad-spectrum claim and a UVA-inadequate product.

One clinical reference worth citing: a 2019 in vitro comparative study (n=24 formulations, tested across 3 independent labs) found that in vitro SPF values overestimated in vivo results by an average of 18% when film application uniformity was not controlled. The coefficient of variation for in vitro SPF dropped from 22% to 9% when a standardized film applicator was used. That’s the kind of method sensitivity that determines whether your SPF 50 claim survives a regulatory audit.

For brands targeting both EU and Asian markets, we recommend running both ISO 24444 in vivo (for SPF labeling) and in vitro UVA-PF (for PA+++ and broad-spectrum claims). The SCCS Scientific Opinion on UV filters provides the safety and efficacy framework that underpins EU filter approvals, and it’s worth reading if you’re evaluating novel UV actives. For China registration, the NMPA Cosmetic Regulation requires domestic testing through an NMPA-recognized CRO — your overseas test data won’t substitute.

See our deeper technical breakdown of mineral UV filter systems and formulation strategy for how we approach ZnO and TiO₂ selection across different SPF targets.

Stability and Compatibility: Where SPF Claims Actually Fail #

This is usually where projects go sideways. A formula can pass SPF testing at T=0 and fail at T=3 months accelerated stability. We’ve seen it more than once. The SPF claim is only as good as the UV filter stability under real storage conditions, and mineral filters behave very differently from chemical filters in this regard.

Zinc oxide and titanium dioxide are photostable — that’s one of their core advantages. But they’re not inert. At pH below 5.5, ZnO begins to dissolve, releasing zinc ions that can destabilize the emulsion and interact with preservative systems. We keep our mineral sunscreen emulsions at pH 6.0–7.0. Below that range, we’ve observed viscosity drop and phase separation within 8 weeks at 40°C/75% RH accelerated conditions. Above pH 7.5, some emulsifiers start to saponify. The window is real and it’s not wide.

Temperature is the other lever. Mineral UV filters are physically dispersed, not dissolved, so their stability is tied to the emulsion matrix. If your emulsion breaks, your SPF breaks with it. We run stability at three conditions: 25°C/60% RH (real-time), 40°C/75% RH (accelerated per ICH Stability Guidelines), and freeze-thaw cycling (5 cycles, -10°C to 25°C). Freeze-thaw is where mineral emulsions fail most often. Particle agglomeration during freezing is hard to reverse, and once ZnO or TiO₂ aggregates, the UV absorbance profile shifts and SPF drops.

Honestly, most brands underestimate the packaging contribution to SPF stability. UV-transparent packaging — clear glass, certain PET bottles — allows photodegradation of co-formulants even when the mineral filters themselves are stable. We’ve had a formula where the mineral SPF held but the antioxidant co-actives degraded enough to change the product color and odor by week 12. The SPF number survived. The product didn’t.

Incompatible combinations are worth calling out directly. High-load niacinamide (>5%) in the same phase as ZnO can form zinc-niacinamide complexes that yellow the formula and reduce UV filter dispersion efficiency. We’ve seen this at 3% niacinamide in some formulations — it depends on the ZnO surface coating. Vitamin C (L-ascorbic acid) at pH below 3.5 is essentially incompatible with any mineral emulsion system; the pH alone destroys the emulsion. And fragrance loads above 0.8% in mineral emulsions are a consistent failure point for us — the solvent effect on the emulsifier system is real.

For brands working on acid exfoliation and UV protection combinations, the pH conflict between AHA systems and mineral emulsions is something we navigate on almost every brief. It’s not impossible, but it requires a two-product system or very careful pH zoning within a single formula.

Where Most Brands Get the Testing Strategy Wrong #

The most common mistake we see: brands submit a formula for SPF testing before stability is confirmed. They get a passing SPF result at T=0, print packaging, then discover at T=3 months that the formula has drifted. Now they have packaging inventory and a formula that needs adjustment. We almost always push back on this brief sequence.

Our standard recommendation is to run a 4-week accelerated stability screen before committing to SPF testing. At 40°C/75% RH for 4 weeks, you’ll catch most emulsion instability, pH drift, and viscosity failures. It’s not a substitute for full 12-week accelerated data, but it filters out the obvious failures before you spend on in vivo panels.

The other thing brands consistently underestimate is the cost of re-testing. A full ISO 24444 in vivo panel at a qualified CRO runs roughly $3,000–$6,000 USD depending on SPF level and number of subjects. If your formula fails stability and needs reformulation, that’s another full test cycle. Airless pump packaging — which genuinely extends mineral sunscreen stability by limiting oxidation and contamination — adds $0.40–$0.80 per unit. At MOQ 3,000 units, most indie brands absorb that. At MOQ 1,000, it’s a harder conversation.

We’re still not fully convinced that in vitro SPF methods will replace in vivo for primary labeling claims in the near term, despite the animal welfare pressure pushing in that direction. The correlation data is improving, but the regulatory acceptance is uneven across markets. What’s acceptable in the EU today may not be accepted by FDA or NMPA tomorrow. This is still evolving.

Formulation Notes for Brand Partners #

What market? What are you expecting on-pack? Those are the first two questions we ask when a sunscreen brief comes in. The answers determine everything — test method, filter selection, pH target, and packaging spec.

If you’re targeting EU + UK with a broad-spectrum claim, you need ISO 24444 in vivo for SPF and in vitro UVA-PF for the PA or broad-spectrum designation. Budget 10–14 weeks for testing plus stability confirmation before launch. If you’re adding China to the scope, add another 6–9 months for NMPA registration and domestic CRO testing.

For mineral-only formulas, we default to a ZnO/TiO₂ combination at a minimum 8% combined load to reliably reach SPF 30. SPF 50 typically requires 15–20% combined mineral load, which creates texture and whitening challenges that need to be solved at the dispersion stage, not the formulation stage. Surface-coated particles help. Nano-grade particles help more, but trigger additional regulatory disclosure requirements in the EU.

If you’re combining actives — niacinamide, peptides, antioxidants — tell us upfront. Compatibility screening adds 3–4 weeks but saves reformulation cycles later. We’ve stopped assuming compatibility and now require a compatibility brief before we finalize the active phase. One pilot batch failed because a brand didn’t disclose a fragrance blend that contained a UV-absorbing component — it shifted the SPF result by nearly 4 points. We now require full INCI disclosure of all co-actives before testing.

It’s not a perfect system. But it’s the one that gets products to market without surprises.

Frequently Asked Questions #

Q: We want to label SPF 50+ — do we need to test at SPF 50 or higher?
For EU labeling, SPF 50+ is a category, not a precise number — any formula testing at SPF 50 or above qualifies. Under ISO 24444, you test to the actual result and label to the nearest category (6, 10, 15, 20, 25, 30, 50, 50+). You don’t need to hit SPF 60 to claim 50+, but your in vivo result must be ≥50 with the lower confidence interval clearing the threshold.

Q: Can we use in vitro SPF data for our US FDA submission?
Not as primary labeling support. FDA treats sunscreens as OTC drugs and requires the FDA monograph test method, which is distinct from ISO 24444. In vitro data can support internal development decisions, but your US SPF claim needs FDA-method in vivo data. Budget separately for this if you’re targeting the US market.

Q: Our formula has 12% ZnO — why did we only get SPF 28 in vitro?
In vitro SPF is highly sensitive to film uniformity and particle dispersion. At 12% ZnO, poor dispersion or agglomeration will reduce effective UV coverage and underperform the theoretical SPF. We’d look at your dispersion process first — specifically milling time and temperature — before adjusting the ZnO load. A D90 above 800 nm is usually the culprit.

Q: How long does full SPF stability testing take before we can launch?
Minimum 12 weeks accelerated (40°C/75% RH) plus real-time initiation. Add 4–6 weeks for ISO 24444 in vivo testing and 2–3 weeks for in vitro UVA-PF. Realistically, from formula finalization to stability-confirmed SPF data, you’re looking at 16–20 weeks. Brands that try to compress this timeline are the ones who end up with packaging inventory and a reformulation problem.

Q: We want PA++++ — what UVA-PF do we actually need to hit?
PA++++ requires UVA-PF ≥ 16 under the JCIA method. In our lab, reliably hitting UVA-PF 16 with mineral-only systems requires a ZnO load of at least 12–15%, combined with a TiO₂ component to fill the UVB range. Without ZnO as the primary UVA absorber, PA++++ is very difficult to achieve in a mineral-only formula. Chemical UVA filters like Tinosorb M or Uvinul A Plus get there more efficiently, but those come with their own EU regulatory status considerations.

Have a product concept in mind? Contact our formulation team to request a complimentary brief review.