The phrase is used for at least four different things, and they do not carry the same risk.
Skin that simply stings. No visible redness, no diagnosis, but acids, alcohol, fragrance and sometimes water sting on contact. This is the most peelable version of sensitive skin. It usually reflects a low irritation threshold rather than active disease, and a low-concentration superficial peel with a single coat is a reasonable trial.
A damaged barrier. Skin that is flaking, tight, shiny in patches, and reacting to products that used to be fine, usually after over-exfoliating, a strong retinoid, or a course of something drying. This is temporary and it is fixable, and the sensible move is to repair it first. A peel on a stripped barrier goes deeper than the same peel on intact skin, because the outer layer that would normally slow the acid down is not there.
Rosacea. Flushing, persistent redness across the central face, visible small vessels, sometimes bumps and pustules. Rosacea skin is reactive by definition, and there is no trial evidence at all about peeling it. Peels are used in rosacea in practice by some clinicians, cautiously, at low strength. That is practice, not evidence.
Eczema, or atopic dermatitis. An inflammatory disease with a genuinely defective barrier. There is no trial of peeling eczema either. Active eczema is the clearest case for waiting: the barrier is not intact, the skin is already inflamed, and a controlled chemical injury on top of an uncontrolled one is not controlled anymore.
Two more situations belong in this list even though people do not always call them sensitivity. Deeply pigmented skin is not more fragile, but it is more likely to respond to any inflammation with dark patches afterward, which changes the calculation about depth. And skin that has just had something else done to it — waxing, threading, a laser, a strong retinoid, a course of another peel — is temporarily sensitive whatever it is like normally.
The gentleness ranking that circulates online is roughly right and its evidence is thinner than it sounds. Here is what has actually been measured.
Mandelic acid. A larger molecule than glycolic acid, so it penetrates more slowly and more evenly. In a randomized trial of 50 people with mild to moderate acne, 45% mandelic acid and 30% salicylic acid over six sessions produced no significant overall difference in acne score, with salicylic acid better for blackheads and whiteheads, mandelic better for inflamed spots, and fewer adverse effects with mandelic. That is a genuine tolerability result at equal efficacy.
Salicylic-mandelic combinations. In a three-arm randomized trial of 90 people with melasma, a 20% salicylic and 10% mandelic combination reduced the melasma score by 60.98 percent against 62.36 percent for 35% glycolic acid, a difference that was not statistically significant. The authors' conclusion was that salicylic-mandelic peels are better tolerated and more suitable for Indian skin. In pooled acne evidence, the salicylic-mandelic combination came out ahead of glycolic acid, at 85.3 percent versus 68.5 percent improvement in total acne score.
Salicylic acid at superficial strengths. The foundational study for "salicylic acid is the safe one in deeply pigmented skin" is a 25-person series in Fitzpatrick types V and VI, using 20% and 30% at two-week intervals after two weeks of priming: 88 percent had moderate to significant improvement and 16 percent had minimal to mild side effects. Twenty-five people. Worth knowing, not worth treating as settled. A separate study of 24 people using 30% salicylic acid measured skin color instrumentally and found a lightening effect rather than darkening.
Buffered glycolic acid. The reason this matters is that plain glycolic acid sits at a very low pH, and pH is a large part of why it stings and why it works. In a split-face randomized trial of 20 people with acne, a buffered 50% glycolic acid at pH 3.0 with 0.5% salicylic acid was compared with Jessner's solution. There was no significant difference in lesion count, acne severity or subjective effect, and the buffered glycolic side had fewer side effects. Same result, gentler delivery, which is the cleanest published demonstration that how an acid is formulated matters as much as how much of it there is.
One correction to a popular claim. Mandelic acid is often described as the best acid for pigment as well as the gentlest. On efficacy it has lost head-to-heads: in a comparison of 30% mandelic against 30% lactic acid for dark circles, lactic acid did better, with 50 percent of the lactic group achieving more than 30 percent improvement and higher satisfaction. Gentle and best are not the same axis.
And the caveat that applies to all of it. Every one of these trials was run in people with acne or melasma and otherwise normal skin barriers, mostly in South Asian populations, with 20 to 90 participants each. None was run in rosacea or eczema. The tolerability findings are real, and applying them to reactive skin is an extrapolation.
This is the single most useful thing to understand before you sit in the chair.
Peels are classified by the depth they reach, not by what is in them. A superficial peel causes a localized injury inside the epidermis, the outer compartment of the skin. A medium peel goes through the epidermis into the papillary dermis just below. A deep peel reaches the mid-reticular dermis. Risk rises steeply with each step, and for sensitive skin, so does the chance of a long red aftermath.
Depth is judged during the appointment by how white the skin turns, not by what was in the bottle. Stringy or patchy light frosting is the superficial endpoint. A uniform white coating with a little redness showing through is the medium endpoint. A solid white enamel look is the deep endpoint. A clinician watching for frost is controlling depth in real time.
That is why "a 35% TCA peel" is not a fixed strength. The same solution reaches different depths depending on how many coats are applied, how thoroughly the skin was degreased first, how thick the skin is in that area, and how hard the applicator is pressed. Published sources do not even agree on where 35% TCA sits: one standard reference calls it medium, another calls medium 35 to 50 percent. Both agree that above 50 percent is deep.
Two practical consequences for reactive skin. First, one coat of a light agent is a genuinely different procedure from three coats of the same agent, and asking how many coats are planned is a fair question. Second, published micron depths for peels are widely repeated but have no primary source behind them, so anyone quoting a precise depth in millimeters is quoting a textbook figure that cannot be traced.
There is also something you may not be able to find out. Professional-use-only products are legally exempt from the requirement to carry a consumer ingredient list. That is why several well-known branded peels publish their ingredients but not their concentrations, and it is why an independent depth classification of those products is not possible from public information. If a clinic tells you a branded peel is medium-depth, that is the marketing description, not a verified fact.
Patch test first if your skin reacts to things. Apply the same product, at the same strength, to a small discreet area — behind the ear or along the jawline is usual — and wait. This is standard practice rather than a trial-tested protocol, and it will not predict everything, but it will catch a frank allergic or severe irritant reaction before it happens across your whole face. It is particularly worth doing where the peel contains an ingredient people react to. Some professional peels contain phenol, hydroquinone, resorcinol or salicylates, and allergy to aspirin is a genuine contraindication to salicylic acid.
Repair the barrier before you peel it. If your skin is currently flaking, stinging in the shower or reacting to a moisturizer, that is not the week. A bland routine, no acids, no retinoid, no scrubs, and a plain moisturizer for two to four weeks changes what the peel will do, because an intact outer layer is what slows the acid down.
Priming is a real thing and it is in nearly every successful protocol. In the studies that produced good results in deeply pigmented skin, patients were prepared for two to four weeks beforehand with a lightening cream, often with a retinoid alongside. Priming is standard practice in those protocols rather than a separately tested intervention, but the pattern is consistent enough to be worth asking about.
Start light and go up, never the reverse. Depth can always be added at the next session. It cannot be taken back.
Stop the actives before and after. Retinoids, exfoliating acids, scrubs and anything described as brightening will all sensitize skin. Your provider should tell you when to stop them; if they do not, ask.
And treat sun protection as part of the procedure, not an afterthought. FDA documented that after four weeks of using alpha hydroxy acids, volunteers showed an 18 percent increase in sensitivity to skin reddening from UV and roughly double the sensitivity to UV-induced cellular damage, and that this reverses after stopping. FDA's own recommended wording for AHA products advises sun protection during use and for a week afterwards. That is the best-sourced aftercare instruction in this entire subject.
Peels are useful, popular and often reasonable. They are also not the only option, and for some skin the correct advice is to do something else.
Do not peel skin that is actively inflamed or broken. Active infection, open wounds and active eczema or rosacea flares are standard contraindications, and they are not judgment calls. A peel is a controlled injury, and control requires starting from intact skin.
Do not peel deeply if your skin is deeply pigmented. Published guidance is graded by depth: superficial peels are frequently used in Fitzpatrick types IV to VI and give good satisfaction, medium peels require caution, and deep peels should be avoided altogether. The reason is dark patches afterward, which is the commonest complication of TCA peeling. Real numbers exist and they are modest, not alarming: 12.5 percent temporary hyperpigmentation with no scarring in a series of 40 mostly type V patients on a strong sequential protocol, and 28 percent at four weeks after a single 15% TCA peel in 18 Korean women, resolved in all but one by 12 weeks.
Be aware how thin the evidence is for peeling skin of color at all. A scoping review screened 473 studies and found 7 that met inclusion criteria. Five were about acne or acne scarring and reported positive results with minimal adverse effects. One reported chemical burns from improper use. Seven studies is the entire base.
And know that a peel may not beat what you would be prescribed anyway. In acne scarring in Fitzpatrick types IV to VI, a randomized trial of 60 people found microneedling produced improvement in 73 percent against 33 percent for a 35% glycolic peel. In a review of the retinoic acid peel, the authors concluded that a properly designed comparison against simply prescribing tretinoin cream has still not been done. If your skin is reactive, an approach that does not involve a chemical injury at all is a legitimate answer rather than a consolation prize.
If you get cold sores, say so when you book, not on the day of the appointment. Antiviral medicine is started in advance, because reactivation happens while the skin is healing and the drug needs to already be working. In a series of 181 patients having perioral phenol peels or dermabrasion, those with a history of cold sores who got no antiviral had an outbreak half the time; with standard antiviral cover that fell to 8.3 percent, and after a higher-dose regimen was introduced no further outbreaks were seen. Notably, 6.6 percent of patients with no history of cold sores at all had an outbreak too. Reactivation has been reported after a superficial peel, so a light peel is not exempt. Published protocols start the antiviral either two days before, the day before, or at the latest on the day. None start afterward.
If you take or recently took isotretinoin, say so, and expect a more specific answer than the old rule. The historic instruction to wait six months came from the drug's package insert and three small case series from the mid-1980s. A 2017 systematic review and consensus covering 32 publications and 1,485 procedures found insufficient evidence to justify delaying superficial chemical peels, manual dermabrasion, cutaneous surgery, laser hair removal, and fractional ablative and non-ablative lasers. A separate 2017 task force reached the same conclusion for superficial peels and non-ablative lasers, and added that superficial and focal dermabrasion may be safe when done by a well-trained clinician. Both stop short of medium and deep peels, which are not covered by the "insufficient evidence to delay" finding. A third consensus from India goes further and includes medium-depth peels and microdermabrasion. The consensus is also not universally accepted and has drawn published criticism. What that adds up to: a superficial peel while on or recently off isotretinoin is supported by current consensus, a medium or deep peel is not.
Tell your provider about aspirin or salicylate allergy, about allergy to phenol, hydroquinone or resorcinol, about any autoimmune or liver condition, about a tendency to form thick raised scars, and about any current skin infection. Ask what the acid is, at what concentration, at what pH, and how many coats are planned. Ask what they would do if your skin frosts faster than expected.
Ask who is performing it and what they are licensed to do. In the states that have written it down, the legal test is depth. Ohio, Texas and California all limit estheticians to exfoliation of the epidermis, and medium and deep peels are treated as medical procedures. Ohio adds a numeric safe harbour — 30 percent concentration and a pH of at least 3, with stronger products allowed only where the manufacturer documents that they do not penetrate below the stratum corneum. There is no federal rule and no reliable nationwide table, so check your own state board.
Do not buy a peel online to do yourself. This is the situation FDA's 2024 warning was written about, and its description of the harms is worth reading in its own words: these products "remove layers of skin to varying depths and may cause severe chemical burns, pain, swelling, infection, skin color changes, and disfiguring scars. These injuries may even require emergency care or specialty care from a dermatologist or surgeon."