Benzene in Aerosol Cosmetics: What the Recall Wave Reveals About Your Testing Protocol

In 2022, Procter & Gamble recalled approximately 30 aerosol hair care products — Pantene, Herbal Essences, Aussie, and Old Spice SKUs included — after internal testing found benzene contamination. Unilever followed within weeks, pulling Suave, TRESemmé, Nexxus, and Dove aerosol dry shampoos. By the time the recall wave crested, more than 100 aerosol cosmetic and personal care SKUs had been voluntarily removed from commerce in a roughly 12-month window. Some of those products contained benzene at levels exceeding 6 ppm. FDA’s interim limit for benzene in drug products is 2 ppm.

That gap — between what was found and what’s acceptable — is a testing problem. Specifically, it’s a problem with where in the manufacturing chain brands were looking, and what analytical method they were using when they looked at all.

How Benzene Gets Into Aerosol Products (It’s Not the Formula)

The assumption most brands carry into their first recall conversation is that benzene must have come from a contaminated active ingredient or a mistake in formulation. It rarely does. The far more common pathway is the propellant.

Aerosol cosmetics — dry shampoos, spray sunscreens, setting sprays, body mists — rely on compressed hydrocarbon propellants, typically isobutane, n-butane, or a proprietary blend. These propellants are derived from petroleum refining, and benzene is a natural impurity in petroleum streams. If the refining process isn’t sufficiently controlled, trace benzene carries through into the finished propellant. From there, it enters every canister filled at your contract manufacturer — without ever touching your formula.

There’s a second pathway that brands rarely account for: benzene can form during storage. Certain fragrance components and preservative precursors can degrade over time into benzene under specific temperature and light conditions. This is less common than propellant contamination, but it explains why finished-product testing done at the time of release doesn’t always predict what a product contains 12 or 18 months later on a retailer’s shelf.

The independent pharmacy testing company Valisure brought this issue into public view in 2021 when it filed a citizen petition with FDA documenting benzene in hand sanitizers and spray sunscreens at levels ranging from trace amounts to above 6 ppm. Valisure’s petition didn’t just trigger recalls — it exposed how systematically most aerosol brands had failed to include benzene in their routine testing panels. That omission had been there for years. The recalls were the consequence.

What Most Cosmetic Testing Protocols Miss

When new cosmetic clients bring us their existing testing documentation, benzene is almost never on the panel. Brands routinely test for heavy metals, microbial limits, pH, viscosity, and active ingredient potency. If they’re exporting to the EU, they may test for regulated fragrance allergens. But residual solvents — and benzene specifically — typically don’t appear until something has already gone wrong.

Part of the reason is categorical. Benzene is a Class 1 residual solvent under USP <467>, the chapter that governs residual solvent testing in pharmaceutical products, with a limit of 2 ppm. But aerosol cosmetics aren’t pharmaceuticals — unless they contain an OTC drug active. Spray sunscreens that carry SPF ratings are regulated as both a cosmetic and an OTC drug under FDA rules. That dual classification creates a systematic compliance gap: brands treating their spray sunscreen purely as a cosmetic may not be applying the USP <467> framework at all, even though the product legally requires it.

For pure cosmetics — a dry shampoo with no drug actives, for example — there is no explicit FDA benzene concentration limit codified in the CFR. But that doesn’t mean benzene is acceptable. Under 21 CFR §601 and FDA’s broader authority over adulterated cosmetics, benzene is not an approved or permitted cosmetic ingredient, and its presence at detectable levels is grounds for an adulteration finding. The regulatory exposure is real whether or not there’s a specific number to point to.

California adds another layer. Under the California Safe Cosmetics Act, any product containing a chemical on the state’s Carcinogen List — benzene has been on it since the program’s inception — must be reported to the California Safe Cosmetics Program if the chemical is present above 0.001% (10 ppm) when intentionally added. Contamination at manufacturing is generally treated as unintentional, which changes the reporting calculus, but it doesn’t eliminate the obligation to know your benzene levels. Brands selling into California’s $7.1 billion personal care market who can’t produce that data are operating on assumptions they shouldn’t be making.

The Analytical Method That Actually Catches It

Headspace gas chromatography-mass spectrometry — HS-GC-MS — is the validated method for detecting benzene in aerosol cosmetics and personal care products. The headspace approach is particularly well-suited to this application because it isolates volatile compounds from the product matrix — propellant, emollients, surfactants — without the matrix interference that would compromise a direct injection approach. USP <467> specifies headspace GC as the procedure for Class 1 and Class 2 residual solvents, and it’s what any credentialed cosmetic testing laboratory should be running.

Detection limits matter here in a way they don’t for some other contaminants. Benzene GC-MS methods in a properly equipped lab should achieve detection limits well below 0.5 ppm — which gives you a meaningful margin before the 2 ppm threshold. If a lab can only report results to 1 ppm or above, that’s insufficient sensitivity for a contaminant with this regulatory history. Ask for the method detection limit (MDL) alongside every result, not just the concentration or a pass/fail notation.

The other technical point worth flagging: because aerosol products are pressurized, sample preparation for headspace analysis requires specific handling to avoid losing volatile analytes before analysis. Labs that primarily run non-pressurized cosmetics may not have optimized their aerosol sample protocol. It’s worth asking specifically whether the lab’s method has been validated for pressurized aerosol matrices, not just liquid or solid cosmetic forms.

Three Sampling Points Your Protocol Needs to Cover

A single finished-product test at release is better than nothing, but it’s not a benzene risk management program. A complete protocol covers three distinct points:

Incoming propellant lots. Each new lot of isobutane or n-butane from your contract filler’s supplier should carry a certificate of analysis with a specific benzene result — not just a “conforms to grade” notation. If your contract filler can’t provide a COA with a method-identified benzene value, or if they’re relying on the propellant supplier’s own data without third-party verification, that’s your first gap to close. The propellant is where contamination enters; it’s where your monitoring should start.

Finished product at release. Testing of the filled, pressurized canister using HS-GC-MS before the product ships. Pull samples from multiple lots when possible — a single lot result, even if clean, doesn’t characterize your ongoing supply chain risk. Three lots gives you a baseline; more gives you a trend.

Stability samples at accelerated conditions. Aerosol cosmetics typically carry shelf lives of 2 to 3 years. A benzene formation event from fragrance or preservative degradation that begins at month 12 means contaminated product has been in commerce for nearly a year before a complaint surfaces. Stability testing at 40°C/75% RH with benzene included in the analytical panel at 3-month and 6-month intervals is your early warning system for the formation pathway.

Running all three points costs more than testing finished product alone. The math becomes obvious, though, after one recall. P&G’s 2022 aerosol voluntary recall involved the destruction of millions of units, plus freight costs, retailer credits, regulatory correspondence, and brand equity damage that doesn’t appear in any single line item.

The Regulatory Direction Is Not Getting More Permissive

FDA hasn’t yet published a formal cosmetic-specific benzene limit, but its enforcement posture since the 2021–2022 recall wave is unmistakable: detectable benzene in aerosol cosmetics is a problem the agency will act on. The Modernization of Cosmetics Regulation Act (MoCRA), signed in December 2022, gives FDA expanded authority to mandate recalls, access facility records, and require safety substantiation files. Brands that can’t produce testing documentation demonstrating active benzene monitoring are in a materially weaker position than they were four years ago, regardless of whether a specific limit exists in the CFR.

The FTC has also been watching claims in the aerosol personal care space. A dry shampoo marketed with “clean,” “safe,” or “non-toxic” positioning that is later found to contain benzene at 3 ppm creates simultaneous problems: an FDA adulteration issue and a potential FTC substantiation failure. Managing those two regulatory problems together is considerably harder than preventing either one.

For brands selling through Amazon, the exposure compounds quickly. CPSC or FDA recall announcements typically trigger listing suspensions within 24 to 48 hours. Reinstatement without documentation takes months — and by then, the listing’s ranking history, reviews, and ad performance have often degraded significantly.

Where to Start if You Don’t Have Current Data

Request the certificate of analysis on the propellant lot used in your most recent production run from your contract filler. Confirm it includes a benzene-specific result with the analytical method identified. If it doesn’t exist or only shows “conforms,” that’s the gap to address first.

Then test your current finished product inventory. Pull samples from three different production lots and submit them to an ISO 17025 accredited cosmetic testing laboratory for headspace GC-MS analysis. Ask for method detection limits, not just results. If you produce any product with an SPF claim, apply the USP <467> framework and treat 2 ppm as your ceiling regardless of whether a specific cosmetic limit exists.

Finally, add benzene to your ongoing stability protocol. Products already in market with 18 or 24 months of remaining shelf life are the ones that matter most — that’s the window where formation events would go undetected without proactive testing.

The brands that came through the 2021–2022 recall period with the least damage had one thing in common: testing data. Not because their data was perfect, but because they could demonstrate a functioning, documented risk management system. That’s the standard FDA, major retailers, and marketplace platforms are now applying to everyone in the aerosol category.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• MoCRA Documentation Strategy: What FDA Expects in a Cosmetic Safety File — Aurora TIC covers FDA regulatory consulting and safety substantiation strategy for cosmetic brands operating under the new MoCRA framework.
• Raw Material Supplier Qualification: COA Review and Incoming Testing for Cosmetic Ingredients — Ayah Labs on building an ingredient-level testing and supplier qualification program that catches contamination before it reaches your formula.
• Canadian Cosmetic Regulations and Health Canada Notification Requirements — Androxa’s team explains how Canadian cosmetic notification and ingredient restrictions compare to the US post-MoCRA landscape for brands expanding north.