Is your hair falling out after a perm — or breaking?

A perm works by rebuilding the internal bonds that hold a hair's shape, and that process unavoidably leaves the fibre weaker and rougher than before. The follicle, where hair is actually made, is untouched. How much damage happens next is partly in your control.

"My hair is falling out after my perm"

It's common to notice more broken pieces on the comb and on the shoulder, and a sense that hair is falling or thinner than before the appointment. The broken pieces look like hair loss because they show up the same way loss does — our piece on Hair Loss explains the difference between breakage (what you experience here) vs hair loss.

Every perm available today works the same way: taking the hair strand apart, chemically, and reassembling it around a new shape. This restructuring leaves it more prone to fracture under ordinary combing and brushing.

The cortex is the layer that gives a strand its strength. Inside the cortex, protein chains are cross-linked by disulfide bonds: strong, sulfur-based bonds that act like rungs holding a ladder's two sides in place. A perm works in three steps: a reducing agent breaks those rungs apart, the hair is wound onto a rod and held in its new shape, and a neutralising agent (an oxidiser) re-forms the rungs in that new position ^[1][2]. The result is a real structural change to the non-living strand.

That's also why a perm lasts until it's cut or grown out, unlike a blow-dry or a wet set, which only rearrange weaker, water-reformable bonds and fall out again in humidity ^[1].

In one sentence: a perm rearranges the internal architecture of the hair strand.

Why that shows up as breakage

Breaking and re-forming disulfide bonds doesn't return the cortex to its original strength. Some keratin protein is lost in the process, and the bonds that re-form don't perfectly recreate the original structure.

Dermatologist Zoe Draelos, writing in a widely used cosmetic-dermatology reference, cites permed hair as 15% weaker than unpermed hair ^[1]. More recent laboratory work on repeated digital perms (which add heat to the chemical process) found protein loss reaching roughly 58% after three treatments with a thioglycolate-based lotion — the same reducing-agent family used in the classic alkaline perm in the table below ^[3].

Both figures point to the same mechanism: a weakened cortex fractures more easily under the everyday stress of combing, brushing and tying hair up — producing more broken ends, not root loss.

The other things you notice — same cause

A perm produces a cluster of changes that can look unrelated but trace back to the same two things: the disulfide-bond rework in the cortex, and the loss of a surface lipid called 18-MEA (more on that below).

The rotten-egg smell — take it seriously

A sulphurous, rotten-egg smell that lingers or reappears after a shampoo usually means the neutralisation (second step of perm) was incomplete — some bonds were never reformed. That's more than an odour problem: un-rejoined bonds mean extra weakness on top of what the perm already causes, so it's worth asking the salon to re-neutralise the hair rather than living with it.

What's changed since the salon rulebook

Older cosmetic-dermatology references describe perm damage as simply unavoidable — the cost of the chemistry, full stop ^[1]. That's still broadly true, but it understates how much control there now is over how much damage happens. The more accurate framing is that the damage is reducible, not eliminable — a combination of gentler chemistries, more careful technique and newer bond-repair actives has narrowed, without erasing, the gap between "curl" and "damaged hair." Those newer approaches also vary considerably in how well they've been tested. Before looking at what helps, it's worth understanding the strength of the evidence behind each claim.

Perm types compared

Three things worth being precise about. First, curl tightness comes from rod size, not chemical strength. A smaller rod gives a tighter curl regardless of which reducing agent is used. Second, chemical strength is tuned to hair resistance. Coarse, resistant or virgin hair typically needs a higher concentration to fully break the bonds, while fine, fragile or already-bleached hair needs less, to avoid over-processing ^[15] — so a "stronger" formulation isn't about achieving a tighter curl, it's about higher power to break the bonds. Third, heat is not automatically gentler. Because digital perms use milder base chemistry, "hot" perms have a reputation for being kinder to hair — but the heat itself is a stressor, and the one study to quantify repeated digital-perm protein loss found a bigger effect than the classic alkaline figure, not a smaller one ^[3]. Hair's internal structure starts to change under sustained heat from roughly 130°C, with more serious protein damage from around 150–200°C — a range digital-perm rods can reach, so the added heat is a real stressor, not a trivial one ^[14].

What actually helps

Some damage is built into every perm, but several decisions can meaningfully reduce how much additional damage occurs afterward. The most useful way to think about this is what a good stylist manages, and what's worth asking for.

Reduce the damage going in. Over-processing — leaving the chemical on too long, using too strong a formulation for the hair's condition — compounds the basic damage a perm always causes. A good stylist will match the chemistry and processing time to the hair in front of them, and will generally avoid stacking a perm directly onto freshly bleached or lightened hair, since that combination compounds damage considerably ^[1].

Sequence chemical services deliberately. Permed hair is more swollen and porous immediately afterward, which means it grabs colour faster and often darker or less predictably than untreated hair. The common professional approach is to perm first and colour after, with roughly ten days between the two services ^[1].

Ask about bond-builders. This class of in-salon treatments, built around a maleate-based active (bis-aminopropyl diglycol dimaleate), works by forming new chemical bridges across the thiol groups left over from the perm's reducing step — effectively adding synthetic cross-links alongside the hair's own re-formed disulfide bonds ^[5]. The mechanism is real and has been shown in laboratory and manufacturer testing; independent, real-world evidence of how much difference it makes on already-permed hair is thinner, and even people within the bond-building industry have cautioned against over-claiming what it achieves ^[6]. Reasonable to treat as an early-stage, plausible category — not a proven undo button. (Evidence: Level 4–5 — lab and manufacturer testing, not yet an independent clinical result on permed hair specifically.) A related, earlier-stage lead uses an enzyme to bond keratin fragments onto already-damaged hair, showing durable improvement through repeated washing in lab testing ^[12] — same repair logic as bond-builders, but not yet an available salon treatment. (Evidence: Level 5 — lab-demonstrated, not a consumer product.)

Manage the surface. A smoothing conditioner reduces friction and eases combing on rough, permed hair — and that's a surface fix for a surface problem (the lost 18-MEA, a fatty layer that normally keeps the cuticle smooth and water-resistant ^[1]), not a structural repair to the cortex. Newer lab research has chemically bonded a similar fatty molecule onto damaged hair's surface — confirmed to stay attached even after washing — to temporarily restore some of that smoothness, though the native lipid itself can't be regrown on a strand that's already left the follicle; only new growth carries it intact ^[7]. (Evidence: Level 5 — laboratory demonstration, not yet a tested consumer product.) For more on what conditioning can and can't do for weathered, porous hair specifically — which describes permed hair well — see Your Hair Mask Rinses Fast.

Set realistic expectations. Some loosening of curl definition in the first week of washing is a normal part of a perm settling in. And because a perm is a permanent change to the hair that's already there, the only real "recovery" is new hair growing in without that chemistry applied to it — which is a question of time, not treatment: hair grows at roughly 1 cm a month on average, though the exact rate varies by hair type ^[13].

Perms and pregnancy

This comes up often enough to address directly: is it safe to get a perm while pregnant?

The available evidence is reassuring, though it is limited rather than extensive. A review of hair-product safety in pregnancy notes that these products have minimal systemic absorption, and describes two case-control studies — one following 525 pregnant women, another following 5,944 — that found no increased risk of preterm delivery or low birth weight associated with chemical hair-straightening or curling exposure ^[8]. That's a meaningfully reassuring pattern, but it rests on observational, case-control studies rather than large randomised trials — Level 3 on the same evidence scale — so "low risk" is a more accurate description than "no risk."

This evidence is specific to thioglycolate-based perming chemistry, not formaldehyde-releasing keratin-smoothing treatments — a different, separate concern, though the two get confused because both are salon chemical services marketed around reshaping or smoothing hair. If you're pregnant and considering a chemical hair service of any kind, it's worth raising it with your doctor — this article can tell you what the evidence says in general, not what's right for your specific pregnancy.

The bottom line

Everything so far in this article treats one thing as settled: that reshaping hair chemically leaves it weaker. That assumption is now being genuinely challenged — not by marketing claims, but by early, peer-reviewed lab research showing hair coming out of a chemical reshaping process stronger than it went in, not merely less damaged.

The clearest examples: both were properly controlled, statistically tested studies. A light-activated method skips the usual alkaline, thiol-based step entirely, using a vitamin-B2 derivative to build new strengthening bonds instead. In the researchers' own head-to-head test, hair treated this way came out with higher tensile strength than untreated hair, while a standard thioglycolic-acid treatment in the same comparison came out weaker ^[9][10]. A second, independent study used short keratin-derived peptides that reshape hair at a gentle, neutral pH instead of an alkaline or acid extreme; it held up for at least 20 washes and made already-damaged hair measurably stronger ^[11]. That's a different promise from the repair actives covered earlier (bond-builders, enzymatic repair) — those work after the fact, on hair that's already damaged; what's described here would prevent the damage from happening in the first place.

Three honest limits keep this from being a "damage-free perm" today. First, neither has been shown to set a curl the way a rod-wound perm does — both were tested as straightening or strengthening treatments. Second, durability falls short of a perm's bar: the keratin-peptide study reports its effect held for at least 20 washes ^[11], which is a real result, but a perm is expected to last until it's cut or grown out. Third, none of this is available to buy or ask for at a salon, and it's likely years, not months, before any of it is.

So the honest update is this: "perming damages hair" is looking less like a fixed law of chemistry and more like a description of the tools available up to now. On the hair you can actually get permed today, though, nothing in this article changes: a perm is a real, permanent change to the hair you already have — done well, on hair that can take it, that's not a catastrophe, but it is a trade. And it's a trade the follicle has no part in: new hair keeps growing on its own schedule, regardless of what happens to the fibre above the scalp.

Perms weaken the strand, not the follicle. Most of what people interpret as "hair falling out" is breakage from a fibre that's become mechanically weaker. The chemistry can't be undone, but careful handling can limit additional breakage while healthy new hair grows in.

Common questions

Why didn't my curl take, or take evenly?

A few common reasons: hair that's too short to wrap fully around the rod, fast new growth that hasn't been shaped by the perm yet, and older, weathered ends absorbing the lotion faster than fresher hair near the roots or nape.

Does a perm damage hair permanently?

Yes, in the sense that it's a lasting structural change to the hair you have right now — hair is a dead fibre, so there's no biological "healing" of that particular strand. The damage is reducible with technique and chemistry choice, not reversible on the hair you have; real recovery comes from new, unpermed growth. One caveat worth knowing: early lab research has managed to reshape hair using different chemistry that comes out stronger, not weaker, than before it started ^[9][10][11] — see "The bottom line" above. That's not available at a salon yet, and it hasn't been shown to set a curl the way a rod-wound perm does, so it doesn't change the answer for a perm you can get today — but "perming always weakens hair" may not hold indefinitely.

Which type of perm is least damaging?

Gentler reducing chemistries — acid or cysteamine-based, at a lower pH — tend to be easier on the fibre than a classic alkaline perm. But heat-based digital perms add their own stress, so "gentler chemistry" and "gentler overall" aren't always the same thing; it depends on the condition of your hair going in.

Editorial note

This article is educational, not medical advice. It focuses on the cosmetic chemistry of perming and the breakage it can cause — not on hair loss as a medical condition. If you're noticing shedding that goes beyond breakage, a widening part, patches of missing hair, or scalp pain, itching or burning, that's a conversation for a dermatologist, not a styling adjustment. If you are pregnant and have specific questions about any hair treatment, speak with your doctor.

Source notes

1. [1] Draelos ZD. Hair Care: An Illustrated Dermatologic Handbook. Taylor & Francis, 2005. Perm mechanism, sequencing, neutralisation, and the cited strength-loss figure.
2. [2] Wang J. Mechanisms of impairment in hair and scalp induced by hair dyeing and perming and potential interventions. Frontiers in Medicine. 2023;10:1139607.
3. [3] Han M, Chun J, Lee J, Chung C. Effects of permanent waving on changes of protein and physicomorphological properties in human head hair. Journal of the Society of Cosmetic Chemists. 2008;59:203–215. (~58% protein loss after three digital-perm treatments with a thioglycolate-based lotion.)
4. [4] DermNet NZ. Glyceryl monothioglycolate allergy (acid perming solution).
5. [5] El Khatib S, Hammoudi Halat D, Khaled S, Malki A, Alameddine B. Novel Compounds for Hair Repair: Chemical Characterization and In Vitro Analysis of Thiol Cross-Linking Agents. Pharmaceuticals. 2025;18(5):632. — bond-builder (maleate/thiol cross-linking) mechanism.
6. [6] Croda Beauty. Bond Building Myth Busting. (industry source; cited for its caution against over-claiming bond-builder efficacy).
7. [7] Song SH, Park H, Lim B, et al. Biomimetics through bioconjugation of 16-methylheptadecanoic acid to damaged hair for hair barrier recovery. Scientific Reports. 2024;14:27387.
8. [8] Chua-Gocheco A, Bozzo P, Einarson A. Safety of hair products during pregnancy: personal use and occupational exposure. Canadian Family Physician. 2008;54:1386–1388. (cites Blackmore-Prince et al. and Rosenberg et al. case-control studies).
9. [9] Kim SY, et al. Green chemistry method for hair strengthening and setting using visible light-mediated protein crosslinking. Journal of Cleaner Production. 2022;363:132535.
10. [10] Kim JH, Kim SY, Choi J, Lee HJ. Visible light-mediated environmentally friendly and universally applicable green chemistry for hair cross-linking. ACS Sustainable Chemistry & Engineering. 2023;11:10029–10040.
11. [11] Cruz CF, Martins M, Egipto J, Osório H, Ribeiro A, Cavaco-Paulo A. Changing the shape of hair with keratin peptides. RSC Advances. 2017;7:51581–51592. (open access; read in full)
12. [12] Liu Y, Liu J, Xiao J. Enzymatic Crosslinking of Amino Acids Improves the Repair Effect of Keratin on Hair Fibre. Polymers. 2023;15(9):2210.
13. [13] Loussouarn G, et al. Diversity in human hair growth, diameter, colour and shape. An in vivo study on young adults from 24 different ethnic groups observed in the five continents. European Journal of Dermatology. 2016. (Reports ethnicity-specific averages — Caucasian and Asian hair grow faster, African hair slower; "roughly 1 cm/month" here is a rounded, cross-group approximation, not a single measured figure.)
14. [14] McMullen R, Jachowicz J. Thermal degradation of hair. I. Effect of curling irons. Journal of Cosmetic Science. 1998;49(4):223–244. (Curling irons run 130–164°C in this study, and measurable fibre damage — tryptophan breakdown, rising combing force — showed up across that range; the 150–200°C "more serious damage" figure is a widely repeated industry/trade estimate, not confirmed here by a single peer-reviewed source.)
15. [15] Chin CM, et al. Comparative Assessments of New Hair-Straightening Cosmetic Formulations on Wavy Type 2 Hair. Cosmetics. 2024;11(6):222. (straightening-formulation study; shows reducing-agent concentration tuned to hair resistance/type — the same underlying principle applies to perming chemistry, which uses the same reducing agents).