U.S. District Judge Analisa Torres of the Southern District of New York was not having it on September 15, when she threw out a consolidated class-action lawsuit alleging that L’Oréal engaged in deceptive business practices, breach of implied warranty, negligent misrepresentation/omission, and unjust enrichment. Basically, the plaintiffs argued that the CeraVe and La Roche-Posay benzoyl peroxide (BPO) acne products were misbranded and adulterated because L’Oréal failed to disclose the presence of benzene in its BPO products.
Before the L’Oréal ruling, a string of federal judges had substantially or fully dismissed similar benzene claims on the same preemption grounds.
Judge Torres granted L’Oréal’s motion to dismiss with prejudice on the grounds that federal law preempts state-law requirements that would impose additional or different labeling requirements on OTC drugs. In other words, states cannot impose an OTC drug labeling requirement that is “different from or in addition to federal requirements.”
And because Torres dismissed the case with prejudice, the plaintiffs don’t get another shot at fixing their complaint. The suit is dead in the water unless they successfully appeal.
The FDA (Food and Drug Administration) Acne Monograph recognizes benzoyl peroxide as GRASE (Generally Regarded As Safe and Effective) at concentrations of 2.5% to 10% for topical, over-the-counter use, according to the monograph. It does not require a benzene warning. The court also rejected the plaintiffs’ argument that benzene should be disclosed as an inactive ingredient. Under federal law, an inactive ingredient is defined as a component intended for use in manufacturing. However, the plaintiffs alleged that benzene forms through BPO decomposition; they did not allege that L’Oréal intended to use it as an ingredient.
This and other BPO lawsuits followed a 2024 FDA Citizen Petition from Valisure, an independent laboratory whose testing methodologies have been criticized for potentially overstating contamination levels and relying on questionable testing conditions.
In its BPO Citizen Petition, Valisure said its data showed that acne products on the market exceeded the FDA guidance's 2 ppm threshold for benzene by more than 800 times. Critics argued that incubating products at 122°F for 18 days did not replicate typical consumer storage conditions. Valisure maintained that the temperature could occur during distribution and is used in pharmaceutical stability testing.
In response, the FDA explicitly warned that unvalidated third-party testing methods can produce inaccurate results and “much higher reported levels” of benzene than are actually present. The agency conducted an independent study that found that more than 90% of the 95 tested BPO acne products tested had undetectable or extremely low levels of benzene, with only 6 products showing elevated levels.
And even if a consumer were to use one of the six products daily for decades, the cancer risk from benzene was “very low.”
The plaintiffs also argued that L’Oréal’s BPO products were “adulterated” because they contained more than 2 parts per million (ppm) of benzene. But the court found no provision in the FDCA (Federal Food, Drug, and Cosmetic Act) or FDA regulation establishing a 2 ppm benzene limit for BPO acne products.
The plaintiffs’ final problem was that they were trying to turn FDA guidance into federal law. They pointed to the agency’s 2 ppm conditional threshold for benzene and argued that a CeraVe product testing above it violated the FDCA. But the FDA guidance concerns benzene used in drug manufacturing, not benzene that may form through degradation over time.
Torres wasn’t buying it. The 2 ppm threshold comes from nonbinding FDA guidance, not the Acne Monograph, the FDCA, or a binding regulation governing BPO acne products. And an FDA announcement about voluntary recalls does not create one.
That distinction reaches beyond benzoyl peroxide. FDA guidance can influence formulation decisions, manufacturing standards, retailer requirements, and recalls. But guidance is not regulation. In this case, that difference was enough to sink the plaintiffs’ adulteration argument—and, ultimately, their lawsuit.
The ruling doesn't resolve whether BPO products actually pose a benzene risk. It resolves whether plaintiffs can invoke state consumer-protection laws to impose requirements beyond those of federal rules. Torres's answer is no.
The FDA has already tested the products itself, found more than 90% had undetectable or extremely low benzene levels, and warned that unvalidated third-party testing can produce inflated results. Now, with federal judges repeatedly finding similar state-law benzene claims preempted, plaintiffs’ lawyers are running out of room to turn that scientific dispute into an economic-loss consumer class action. Unless an appeal changes the legal calculus, the benzene-in-BPO class-action wave may have already crested.
