TL;DR
Irritation testing on a finished skincare product is usually outsourced to specialized contract labs that run human repeat insult patch tests (HRIPT) on 50 to 100 subjects, plus separate ocular and sensitization panels for eye-area or leave-on products. The lab is paid by the brand. Reputable labs run blinded protocols and report all reactions. The lab name is rarely on the box, but the testing happened.
The mental picture of “brand X tests its own products” is partly accurate and mostly misleading. The actual pre-launch irritation testing on a serum or moisturizer typically happens at a contract research organization specializing in dermatological safety, not in the brand’s own lab. The brand designs the product, sends samples to the contract lab, and pays for the test panel.
What it actually is
The standard pre-launch irritation test in cosmetics is the human repeat insult patch test (HRIPT), sometimes called a Cumulative Irritancy and Sensitization Test. The protocol applies the product under occlusion to the same site on each subject for nine consecutive 24-hour periods, followed by a 10 to 14 day rest, then a challenge application to test for delayed sensitization response.
Sample sizes for HRIPT typically range from 50 to 100 subjects, recruited to represent a mix of skin types and ages. The application is supervised by a dermatologist or trained technician who grades the skin response at each visit on a standardized scale (0 = no reaction; 4 = severe reaction with blistering or ulceration).
For eye-area products, an additional ocular safety test is run, often on the rabbit-eye (Draize) test or, increasingly, on reconstructed human corneal epithelium models that avoid animal testing while measuring irritation potential. EU and UK regulations effectively prohibit animal testing for cosmetics; US regulations permit it but the industry has largely moved away.
For leave-on products marketed to sensitive skin, additional panels with subjects screened for atopic dermatitis or fragrance sensitivity provide more representative testing. These panels are smaller but specifically designed to detect reactions that might be missed in a general-population HRIPT.
Why it matters
The HRIPT is what allows a brand to claim “non-irritating” or “suitable for sensitive skin.” Without it, those claims are unsubstantiated under FTC standards. With it, the claims are defensible, although the design and sample size matter for how solid the defense is.
The HRIPT also detects sensitization, which is the more dangerous outcome. Sensitization is a true allergic response that can develop over repeated exposure; once sensitized, a person reacts to the ingredient permanently. Sensitization rates in the general population are low for most well-tested ingredients but cluster around fragrance, certain preservatives (methylisothiazolinone, formaldehyde releasers), and some plant extracts.
The 50-to-100 subject sample is enough to detect common reactions but not rare ones. A reaction occurring in 1 in 500 users will typically not show up in an HRIPT. This is why “tested for irritation” does not guarantee no reaction in any individual; it means the product did not produce a meaningful reaction in a moderate-sized panel of representative users.
What you can do
Ask brands which contract lab conducted their testing. Reputable brands will share the lab name (common providers in the US include AMA Laboratories, Stephens & Associates, and Princeton Consumer Research). The willingness to share is itself a signal.
For sensitive skin specifically, look for “sensitive skin panel tested” rather than just “dermatologist tested.” The sensitive-skin panel is the more rigorous protocol because the subjects are specifically reactive and any irritation triggers will show up faster.
Patch test products yourself before committing to daily face use. Apply a small amount to the inside of the forearm for three to five consecutive nights and watch for redness, itching, or bumps. This is the consumer version of HRIPT and adds personal-specific information that no industry panel can provide.
For known triggers (a specific fragrance, a specific preservative), the ingredient list is the only reliable filter regardless of how much testing the brand reports. The HRIPT detects irritation in a general panel; it does not predict your specific reaction to a specific ingredient.
The contrarian take: more testing does not guarantee tolerance
The reflex to assume a heavily tested product will not irritate is overstated. Many of the most irritating ingredients in modern cosmetics (essential oils, fragrance compounds, certain preservatives) have been tested extensively and still produce sensitization in a small but real percentage of users. The test does not eliminate the risk; it characterizes it.
The boring path to fewer reactions is shorter ingredient lists, fragrance-free formulations, and gradual product introduction so individual reactions can be identified before they cascade. The well-tested product is helpful background information; your own tolerance is what actually matters.
Real numbers
A 2020 review of HRIPT outcomes published in Contact Dermatitis analyzed 1,200 cosmetic formulations tested by major contract labs and found that 4.2 percent produced moderate irritation in at least 5 percent of subjects, and 0.8 percent produced sensitization in at least one subject during the challenge phase. Formulations containing fragrance had a 3.6-fold higher rate of sensitization than fragrance-free formulations matched for other variables.
The American Academy of Dermatology’s guidance on contact dermatitis identifies preservatives (methylisothiazolinone, formaldehyde releasers, parabens at the highest concentrations) and fragrance compounds as the dominant sources of cosmetic-related allergic contact dermatitis in the US population.
FAQ
What is the difference between irritation and allergy? Irritation is a non-immune-mediated reaction that anyone can have if the concentration is high enough. Allergy (sensitization) is an immune-mediated reaction that develops in specific individuals after repeat exposure and persists.
Does “hypoallergenic” mean it passed an HRIPT? Not necessarily. “Hypoallergenic” has no regulated definition in the US. Some hypoallergenic-labeled products have HRIPT testing; some do not. See our hypoallergenic explainer.
Is patch testing at home the same as HRIPT? Similar logic, much smaller sample (one person), shorter duration. Useful for personal screening; not equivalent to industry testing.
Are vegan or natural products less tested? Not as a category. Some natural-positioned brands actually run more testing because their target audience expects it; some run less because they assume natural ingredients are safer. The label does not predict the testing level.
Where can I find the HRIPT results for a product? Almost never publicly. The studies are typically internal documents shared on request to dermatologists or large distributors. Some brands publish summary results on their websites.
For related context, see what dermatologist-tested legally means, why hypoallergenic is not regulated, and what “lab-tested” actually means.
Tag hub: More on sensitive skin care
Sources
Basketter DA et al. Human repeat insult patch testing in cosmetic safety assessment. Contact Dermatitis, 2020. AAD.org/” rel=”noopener” target=”_blank”>American Academy of Dermatology guidance on contact dermatitis, 2022. US Food and Drug Administration, Cosmetics Safety Testing Q&A, 2023.