What did the study find?

Of 18 skin products bought from online retailers, 6 contained clobetasol propionate, a strong prescription-only steroid that was not on the label. Two of the six also contained an antifungal drug. The same result came back in three separately bought units of each.

It is a corticosteroid, a drug that calms inflammation, and it belongs to the strongest class of steroids applied to skin. In the United States it is available only on prescription. Approved products contain 0.05%.

Why is a hidden steroid a problem?

A strong steroid can make a rash or dark patch look better, which is why such a product seems to work. Used for long periods without supervision, steroids of this strength can thin the skin and cause acne and color changes, and the face is especially vulnerable.

Does this mean a third of these products contain steroids?

No. The researchers chose products they considered likely to contain hidden drugs. They describe the one-in-three figure as the yield of a targeted search, not the rate across the market.

What should someone using such a product do?

The FDA has said of one such product that stopping abruptly may cause reactions and that users should contact a health professional. What to do about a particular product and the condition underneath is assessed by their doctor. This is general information rather than medical advice.

A potent prescription-only steroid was found, unlabeled, in 6 of 18 skin products bought from online retailers

The products claimed to help eczema, psoriasis or dark patches and listed no proven drug. All six contained clobetasol, in one case at nine times prescription strength. The sample was small and chosen to find problem products.

Some skin creams sold online seem to work remarkably well on eczema, psoriasis or dark patches, even though their labels list no drug. A laboratory analysis offers one explanation. In 6 of 18 such products, researchers found clobetasol propionate, one of the strongest prescription steroids used on skin. It was not on any of the labels.

The study, published on October 8 in JAMA Network Open, was small and was aimed at products the researchers thought likely to be adulterated. It does not show how common the problem is across the market. It does show that the problem exists in widely reviewed products.

What was found in the skin products

The headline result is one drug, found again and again. The superpotent (class I), prescription-only corticosteroid clobetasol propionate was detected in 6 products (33.3%), 2 of which additionally contained the antifungals miconazole or terbinafine. A corticosteroid is a drug that calms inflammation. Steroids for the skin are ranked in classes by strength, and class I is the strongest. Antifungals treat fungal infections such as ringworm.

The amounts varied enormously. Detected concentrations of clobetasol propionate ranged from approximately 0.01% to 0.46%, with the highest concentrations exceeding those in FDA-approved topical formulations (0.05%) by 9.2 fold. FDA is the US Food and Drug Administration, and topical means applied to the skin. The weakest product, then, carried a fifth of the prescription strength, and the strongest about nine times it.

The labels gave no hint. None of these clinically proven active ingredients were disclosed on product labels.

These were not obscure items. The six positive products had more than 10,000 customer reviews between them, and the paper adds that several were ranked among the top-selling items in their respective categories at the time of analysis.

How the skin products were chosen and tested

The background is a large market. Skin conditions affect approximately 85 million individuals in the US alone, a figure that traces to a national report which found that nearly 85 million Americans were seen by a physician for at least 1 skin disease in 2013. Alongside prescription treatment sits a trade in products anyone can buy. Numerous over-the-counter (OTC) skin products marketed online claim benefit for conditions such as psoriasis, eczema, and hyperpigmentation despite lacking clinically proven active ingredients. Hyperpigmentation means patches of darker skin.

What is new here is the method. Hidden drugs in skin products have previously come to light one case at a time, and the authors write that no widely adopted standardized approach exists to systematically identify them. So the researchers, Zaur Abilov, Yury Popov and two colleagues, went shopping. Skin products were purchased from major online retail platforms and selected using predefined criteria. A product had to claim it worked on an inflammatory or persistent skin condition. It had to show an absence of listed clinically proven ingredients for these conditions. And it had to be popular, which the team measured as consumer engagement, defined as 100 or more reviews on at least 1 platform. Products that resembled ones already caught with hidden drugs could be included on fewer criteria.

Eighteen products were bought between February and December 2025. Each was put through analytical screening for 28 corticosteroids and antifungals, using a method that separates the chemicals in a sample and identifies each one by its mass.

A single positive bottle could be a fluke or a bad batch, so the team checked. When a hidden drug turned up, 2 additional units of the same product were purchased and analyzed, resulting in 3 samples evaluated per product. Every positive product was positive in all three.

Why an undeclared steroid matters

A strong steroid will often make inflamed or darkened skin look better. That is the point the authors make about hidden drugs in general. Such ingredients may explain apparent efficacy but may also pose substantial risks when used without medical supervision.

The risks are well documented for steroids used knowingly. A review of the subject lists what goes wrong on the skin. The most frequent adverse effects include atrophy, striae, rosacea, perioral dermatitis, acne, and purpura. Atrophy is thinning of the skin, striae are stretch marks, perioral dermatitis is a rash around the mouth, and purpura is bruise-like discoloration. The same review notes that problems build with time and depend on where the product goes. Cutaneous adverse effects occur regularly with prolonged treatment and are dependent on the chemical nature of the drug, the vehicle, and the location of its application. Cutaneous means of the skin, and the vehicle is the cream or spray that carries the drug.

Location is where the new results are most worrying. Due to increased corticosteroid absorption in facial skin, topical treatment is generally limited to low- to medium-potency corticosteroids to prevent local adverse reactions, while periocular exposure has been associated with glaucoma. Periocular means around the eyes, and glaucoma is damage to the nerve of the eye, usually from raised pressure inside it. Yet the team reports: “We found that 3 skin-lightening products contained superpotent corticosteroid.” Such products are commonly applied to the face and may be used repeatedly or chronically by unsuspecting consumers.

Effects are not confined to the skin. The review records that systemic reactions such as hyperglycemia, glaucoma, and adrenal insufficiency have also been reported to follow topical application. Systemic means affecting the whole body. Hyperglycemia is high blood sugar, and adrenal insufficiency is a shortfall in the body’s own steroid hormones. Children are more prone to the development of systemic reactions to topically applied medication because of their higher ratio of total body surface area to body weight.

Stopping can be a problem too. A systematic review of withdrawal reactions found 34 studies meeting inclusion criteria and concluded that withdrawal from topical corticosteroids (TCS) is a real, if poorly studied, condition. TCS withdrawal was reported mostly on the face and genital area (99.3%) of women (81.0%) primarily in the setting of long-term inappropriate use of potent TCS. Burning and stinging were the most frequently reported symptoms (65.5%) with erythema being the most common sign (92.3%). Erythema is redness. The reviewers rated the evidence as low in quality.

Someone using a mislabeled product has no way to weigh any of this, because they do not know they are using a steroid at all.

A long history of hidden ingredients in skin products

None of this is new, which is part of the authors’ argument. Concerns about undeclared corticosteroids in OTC skin products have been reported for decades, prompting regulatory actions, such as FDA warning letters and safety alerts.

One early report appeared in 1999 in the BMJ, a British medical journal. Doctors analyzed herbal creams that patients had brought to skin clinics. There were 11 creams. Eight creams contained dexamethasone at a mean concentration of 456 micrograms/g, which is about 0.05% by weight. Dexamethasone is another corticosteroid. All were applied to areas of sensitive skin such as face and flexures, flexures being skin folds such as the inside of the elbow.

The pattern has persisted. In October 2024 the FDA issued a public warning about an aerosol spray sold for psoriasis, eczema and dandruff. FDA testing identified a high-potency glucocorticosteroid (immunosuppressive and anti inflammatory drugs) in Skin-Cap aerosol spray products that is not disclosed on the label or packaging. The agency said the spray had been found for sale on large retail websites.

The problem is not limited to online sales. In 2024, dermatologists surveyed 33 beauty supply stores in three US cities after seeing patients with steroid acne. They were looking for creams containing clobetasol or a similar steroid. These products are marketed as fade creams to treat hyperpigmentation and uneven skin tone. The results were lopsided: out of 15 stores visited in Miami, 14 (93.33%) contained skincare products with clobetasol, and none of the stores in the other two cities did. The authors pointed out that steroids of this strength are legally sold only on prescription in the United States.

Why undeclared ingredients keep reaching the market

The explanation lies in how different kinds of product are regulated. The paper puts it in one sentence. FDA-approved drugs are subject to premarket review, but nonprescription products do not undergo comparable testing.

The FDA says the same in its guidance to consumers. Although FDA cannot test these products before they enter the market because of federal regulations, it tests as many products as possible once they are on the market. It is plain about motive: some manufacturers intentionally do not disclose that their products contain drug ingredients to try to get around the agency’s review. And it is plain about the law. Under federal law, these products must accurately list their ingredients on their labels, and they cannot contain hidden drug ingredients.

Enforcement runs behind. The agency keeps a public database of products it has caught, with a caveat attached. This database covers only a small fraction of the contaminated and/or fraudulent products on the market. It adds that if a product is not on the list, that does not mean it is safe.

The agency also notes where such products now turn up. Products contaminated with hidden ingredients are increasingly appearing in popular online marketplaces.

One line of the FDA’s guidance bears directly on how the researchers picked their sample. Beware of products with multiple rave reviews touting the effectiveness of the product, the agency says. The study’s own entry requirement was at least 100 reviews of any kind.

The result, in the authors’ words, is that such products may remain available to consumers until they are identified through targeted surveillance or adverse-event reporting. They read their own results as suggesting that regulatory and surveillance approaches remain insufficient.

What the product study cannot show

How common this is. Eighteen products is a small sample, and it was not a random one. The authors say their results represent a yield of a targeted assessment rather than true US market prevalence. Among all skin products sold online, the share with hidden steroids could be far lower.

What else is in them. The screen looked for 28 named drugs. Anything outside that list would have been missed. The authors raise a subtler gap as well: undisclosed corticosteroids may be present in an alternative ester form that differs from commonly registered pharmaceuticals, and screening generally targets known or registered pharmaceutical compounds. An ester form is a chemical variant of the same drug. For that reason the six may be an undercount.

Whether anyone was harmed. No users were studied. The paper lists among its limitations the lack of data on actual population exposure. It shows what is in the containers, not what happened to the people who used them.

Whether it is still true. The products were bought in 2025. Sellers and formulas change, and a product that tested positive then may not be on sale now, or may be on sale under another name.

On what a user should do, the FDA’s 2024 warning is the nearest official guidance. It said that abruptly stopping the use of this product may cause adverse reactions, and that consumers using this product should contact their health care professional to determine the best course of action.

What to do about a particular skin product, and how to treat the condition underneath, is assessed by their doctor.

What the study shows is that, for products like these, the label cannot be taken as a full account of the contents.