In-chair whitening versus home trays for sensitive teeth: which hurts less?

In-chair whitening and home trays carry about the same risk of tooth sensitivity: a 2025 updated systematic review and meta-analysis that pooled 23 trials found no significant difference between them (RR 0.82, 95% CI 0.61 to 1.10)¹. The same review found the sensitivity less intense at home, by a margin that only just clears zero on low-quality evidence, and in both settings the strength of the peroxide and the time it spends on the tooth change the pain more than the chair or the tray does¹²³. S3 Sensitivity Science™, a daily sensitivity toothpaste rather than a bleaching product, keeps full adult-strength fluoride: 1450 ppm, as sodium monofluorophosphate. No trial has tested it, or any other toothpaste, as a way of making either technique hurt less.

What was checked22 peer-reviewed studies and reviews, UK legislation, a European Commission scientific committee opinion, the NHS, the Oral Health Foundation, the General Dental Council, and ASA and CAP pages

What actually differs between in-chair whitening and home trays?

Mostly the dose: how strong the gel is, how long it stays on the teeth, over how many sittings, and who puts it there. In the chair, the gums are shielded, a strong gel goes onto the teeth, a light is sometimes shone on it, and the appointment lasts about an hour in the Oral Health Foundation's description, or one to two hours for laser whitening in the NHS's¹⁰⁶. At home, the dental team makes trays that fit your teeth and prescribes a weaker gel that you wear for a set time each day or overnight, for one to two weeks according to the Oral Health Foundation and around two to six weeks according to the NHS¹⁰⁶. Either way, the General Dental Council says whitening should follow an assessment by a registered dentist and be done on that dentist's prescription, by the dentist or by a registered hygienist, therapist or clinical dental technician¹¹.

In the trials on this page, the in-chair gels were 35% or 40% hydrogen peroxide, given in one to three sessions, apart from one trial arm at 6%⁴⁸⁹¹². The home gels in those randomised trials were 10% or 16% carbamide peroxide, or 4% to 6% hydrogen peroxide, worn from two half-hour periods a day to longer daily wear, over two or three weeks⁹¹³⁸¹⁴. The trials inside the 2016 review ranged wider: at home, 10% to 32% carbamide peroxide worn two to ten hours a day for six to 28 days, and in the chair, 25% to 38% hydrogen peroxide, according to the protocol table of a 2024 umbrella review¹⁵.

Carbamide peroxide is weaker than its number suggests. It breaks down into urea and hydrogen peroxide on contact with water, and the European Commission's Scientific Committee on Consumer Products, in its 2007 opinion on hydrogen peroxide in tooth whitening products, gave a 10% carbamide peroxide gel as yielding a maximum of 3.6% hydrogen peroxide and a 15% gel as the equivalent of 5.3%¹⁶. This page does not convert the stronger carbamide gels used in some of the reviewed trials. Neither route lightens fillings, crowns or veneers, and the Oral Health Foundation says results can last several years, though this varies¹⁰.

The table sets the techniques side by side: what the trials used, what the NHS and the Oral Health Foundation describe, and who may provide each under UK rules⁷¹¹.

Which hurts less, according to the trials?

Neither technique wins cleanly. The reviews that pool the trials find the risk about equal, the newest finds the pain milder at home, and the single studies split, one of them leaning the other way¹¹⁸¹⁹.

The first pooled answer came in 2016. A systematic review and meta-analysis of randomised trials, 12 in its qualitative analysis and eight pooled, found no significant difference between home and in-office bleaching in the risk of sensitivity, its intensity or the colour change¹⁸. Its authors added that this overall comparison ignored the different gel strengths, daily wear times and session counts of the trials it combined¹⁸.

The 2025 update, with 23 trials in its meta-analysis, kept half of that answer and changed the other half¹. In the updated meta-analysis the risk of sensitivity still did not differ significantly (RR 0.82, 95% CI 0.61 to 1.10), but the intensity was lower with home bleaching (standardised mean difference −0.78, 95% CI −1.53 to −0.03), and colour measured by instrument changed more at home (0.49, 95% CI 0.14 to 0.84) while the shade-guide result did not differ¹. Its authors rated the quality of that evidence low¹.

Look at where that intensity interval ends. Its upper limit, −0.03, sits a whisker from no difference at all, so the review is a lean towards trays on low-quality evidence, not proof that trays are gentle, and nothing in it tells a sensitive reader that either route will be painless¹.

A 2024 umbrella review stepped back and read 28 systematic reviews¹⁵. It found no difference between the techniques in colour change or in sensitivity after treatment, then graded the reviews themselves with AMSTAR 2, a checklist for how well a systematic review was carried out: they ranged from critically low to high, and the 2016 home-versus-office meta-analysis was rated low¹⁵. A corrigendum to the umbrella review appeared later in 2024, and its content could not be read for this page¹⁵.

The most carefully blinded comparison gave everyone a dummy of the treatment they were not having. In a triple-blind randomised trial of 130 volunteers, run with gels donated by their manufacturer, half had 35% hydrogen peroxide in the chair and a dummy home gel, and half had 10% carbamide peroxide at home and a dummy chair session, all after a potassium nitrate desensitising gel²⁰. In that manufacturer-supplied trial the chair group reported more sensitivity on the first day only, a mean score of 1.57 against 1.20 on a five-point scale that starts at one for none, and the home group finished with the larger colour change on some measures; sensitivity was a secondary outcome²⁰.

Outside a trial, the result tipped the other way without reaching significance¹⁹. In a practice-based clinical study of 171 patients in Scandinavian practices and university clinics, 50.3% reported sensitivity after home bleaching and 39.3% after in-office bleaching, a difference that was not statistically significant¹⁹. Patients and dentists chose the technique rather than leaving it to chance, the in-office group was small, and gum irritation was more common in it, 35.7% against 14.0%¹⁹.

Pooled patient data from 11 clinical trials, all run by one research group, leaned the same way as the 2025 review but further: the risk of sensitivity was 51% with home bleaching and 62.9% in the chair, and technique predicted sensitivity along with the starting shade of the teeth²¹. In a four-group randomised trial of 48 participants, home bleaching with 10% carbamide peroxide, in-office bleaching with 40% hydrogen peroxide and two combined sequences whitened to a similar degree, and sensitivity scores were lowest with trays alone⁹.

Several of these sources share authors. Researchers listed as A Reis and AD Loguercio both appear on the first home-versus-office review, the concentration review, the pooled patient analysis, the review of light activation, the combined-course trial and the fluoride-gel trial, and one of the two appears on the 2025 update and on the trial of the lower-strength gel^182211781314. That is not a fault in itself, but it means the sources are less independent of one another than their number suggests.

Nobody in these comparisons was chosen for sensitive teeth

This page's searches in September 2026 turned up no comparison of the two techniques in people recruited because their teeth were already sensitive. A PubMed search pairing in-office or at-home bleaching with dentine hypersensitivity or sensitive teeth, and with randomisation terms, returned nine records, none of them such a comparison; a broader search returned two hundred and sixty, and none of the home-versus-office comparisons among them that this page read describes recruiting for sensitivity. The two trials above that publish their eligibility rules shut such people out: the manufacturer-supplied triple-blind trial excluded anyone presenting tooth sensitivity, abrasion, erosion or gum recession, and the trial of 6% against 35% excluded anyone with a history of sensitivity, receding gums or exposed dentine²⁰⁴. The practice-based study enrolled them, and found that sensitivity at the start predicted side effects¹⁹. The pooled reviews do not say in their abstracts how many of their trials did the same.

Every result in this section and the next two is in one table, with the design, size, certainty and funding each source states.

Does the concentration matter more than the setting?

For how often sensitivity happens, yes, as far as the evidence goes, and the gentler gel also whitened less in the largest recent trial²⁴.

Pooling 25 randomised trials of in-office bleaching, a 2020 meta-analysis found that low or medium hydrogen peroxide strengths cut the risk of sensitivity by a third compared with high strengths (RR 0.67, 95% CI 0.51 to 0.86), on moderate-certainty evidence². Colour did not differ significantly in that review, on low or very low certainty, and its authors write that the ideal concentration is still unknown².

A trial published in 2026 then tested a gel at the 6% figure directly⁴. Three 50-minute sessions of a 6% gel left 44% of 140 adults with some sensitivity in a double-blind randomised trial, against 74% of those given 35% on the same schedule⁴. In that double-blind trial the intensity stayed lower with 6% for the first 24 hours and was no different by 48 hours, while on colour the 35% gel came out ahead on every instrument at every time point, and the test for equivalence failed⁴. The gentler course was the weaker one.

A second 2020 meta-analysis, of seven pooled studies with 649 patients, also reported more sensitivity at 35% and above, but its printed interval runs from 0.44 to 1.03, which includes no effect even though it gives p = 0.04, so this page takes the direction and leaves the number²².

Time behaves the same way at home. A 2024 systematic review and meta-analysis of randomised trials found fewer sensitivity events when the daily wear of a home gel was shortened, while most colour measures favoured wearing it for the full time the maker recommends, and its authors advise caution before cutting wear³.

Set those findings beside the comparison of settings and they fit. An in-chair session puts a strong gel on the teeth for under an hour at a time; trays put a weak gel on them for anything from half an hour to several hours a day, over weeks; and the pooled comparisons, which mix every strength and schedule, find the risk about equal¹⁸¹⁵⁸. The 2016 review says plainly that it did not account for those differences, and the abstract of its 2025 update does not say that it did¹⁸¹. What sensitivity feels like in the days after a course, and how long it usually takes to settle, is covered in the Journal's article on sensitive teeth after whitening.

Does the light or laser make in-chair whitening better or gentler?

Not according to the systematic reviews, though the NHS page says it is more effective, and anyone comparing quotes will meet both views¹⁷⁵⁶.

The NHS page on teeth whitening, last reviewed in August 2025, describes whitening with a laser at a dental surgery as quicker and more effective, and says of it: "It's faster and more effective than using gel on its own"⁶. The Oral Health Foundation says only that a special light may be used to help the process¹⁰.

The reviews tested the light directly. A 2018 systematic review and meta-analysis of 21 randomised trials found that shining a light on in-office gel changed neither the colour result nor the risk or intensity of sensitivity, at low or at high hydrogen peroxide strengths, with moderate certainty for instrument-measured colour and for risk¹⁷. Its authors wrote that the commercial claim that light improves and accelerates colour change was not confirmed¹⁷. A 2026 umbrella review of ten systematic reviews reached the same place: light or laser activation gave no consistent improvement in the final result, did not reduce sensitivity, and came out less favourable for sensitivity with some high-energy protocols⁵. The 2024 umbrella review adds that light did not raise the risk or intensity of sensitivity in the reviews it pooled¹⁵.

That leaves a reader with two authorities that disagree. This page gives more weight to the reviews, because they pool randomised trials and publish how they did it, while the NHS page cites no source for its sentence; that is a judgement about kinds of evidence, and a dentist may have practical reasons of their own for using a light¹⁷⁵⁶. Advertising rules have noticed the same gap: the Committee of Advertising Practice advises that an ad should not suggest a whitening treatment involves lasers when it does not, and that anyone advertising laser teeth whitening should hold evidence for every claim²³.

One use of light did help. The 2026 umbrella review found that photobiomodulation, low-level light aimed at easing pain rather than at bleaching, reduced sensitivity after bleaching in placebo-controlled trials without reducing the colour change⁵. It is a separate procedure, and asking whether a clinic offers it is a different question from asking whether it whitens with a laser.

What do the UK rules mean for reading these trials?

That every in-chair trial arm on this page except one used a gel stronger than the highest concentration UK law sets for a tooth whitening product, while the home gels and that one arm sat within it⁷¹⁶⁴.

The cosmetics regulation as it applies in Great Britain covers hydrogen peroxide, and compounds that release it, carbamide peroxide among them, in entry 12 of its Annex III⁷. Oral products, whitening products included, may contain or release no more than 0.1%⁷. Above that, tooth whitening or bleaching products of up to 6% may be sold only to dental practitioners; in every cycle of use a dental practitioner makes the first use, or supervises it directly, before the product goes to the consumer to finish the cycle, and it is not for anyone under 18⁷.

Read against that entry, several home arms of the trials resemble a UK home course. On the committee's figure a 10% carbamide peroxide gel yields at most 3.6% hydrogen peroxide, and the home hydrogen peroxide gels in the combined-course trials were 4% and 6%¹⁶⁸¹⁴. Most in-chair arms of the randomised trials do not resemble a UK chair session in strength: they used 35% or 40% gels^4891412. The 6% arm of the 2026 trial is the in-chair evidence closest to the UK figure, and it caused less sensitivity and less colour change than 35%⁴.

None of that makes a trial run elsewhere wrong, and this page does not say what gel a UK dentist can or cannot use. This page looked for a lawful route to anything stronger in the two places it would expect to find one written, the Annex III entry and the General Dental Council's patient page, in September 2026, and found none described; that is a statement about those two pages and not legal advice⁷¹¹. Who may whiten teeth, and what happens when someone else does, is on the page on whether tooth whitening is legal in the UK.

What about combined courses, and strips from a shop?

A combined course added speed and, against home trays alone, more sensitivity in the trial that compared them; strips sold in UK shops are a different and far weaker product⁸⁷.

In a single-blind randomised trial of 80 participants, one in-chair session of 35% hydrogen peroxide followed by three weeks of 4% hydrogen peroxide at home was set against the same home course alone⁸. In that randomised trial the combined group had the higher risk of sensitivity (RR 1.4, 95% CI 1.1 to 1.9, which the authors express as a 40% lower risk at home) and more intense sensitivity in the first week, by 2.3 points on a ten-point scale, along with more whitening that week⁸. When the course ended, the two groups' colour no longer differed significantly, though the authors' own conclusion calls the combined result slightly higher⁸.

Compared with in-chair whitening alone, a combined course came out differently. In a randomised trial of 66 subjects, adding a prefilled 6% hydrogen peroxide home tray between in-chair sessions with 35% gel gave more colour change by the third session, with a similar intensity and risk of sensitivity¹⁴. And in the four-group trial, trays first and the chair second produced more sensitivity than trays alone, with no difference in colour⁹. So whether a combined course hurts more depends on what it replaces: in these trials it added sensitivity to a home course and did not add it to a chair course⁸¹⁴.

Strips and kits sold to the public in the UK may contain or release no more than 0.1% hydrogen peroxide, and the Oral Health Foundation says shop products may remove some surface stains but usually leave the natural colour of teeth much as it was⁷¹⁰. What those products contain, and how far their trials apply, is on the page on whitening strips and sensitive teeth.

Is either technique "enamel-safe", and what do "ADA" and "fluoride" mean here?

No UK rule this page read defines "enamel-safe", the ADA seal belongs to an American scheme, and the fluoride tested around bleaching was a professional gel rather than a toothpaste^24231312.

What the laboratory evidence shows is a small effect²⁵. A 2026 systematic review and meta-analysis of laboratory studies, nine pooled from 81 that were eligible, found that peroxide bleaching lowered enamel microhardness by a small amount (ratio of means 0.89, 95% CI 0.84 to 0.94), without clear evidence of a clinically meaningful change²⁵. Every study in it measured enamel in the laboratory, so it says nothing directly about a person's mouth²⁵. Everyday enamel wear is covered in the Journal's article on tooth enamel erosion.

"Enamel-safe" is packaging language. The ASA website's search returned no results for "enamel safe", "enamel-safe" or "safe on enamel" in September 2026, against 22 results for "teeth whitening", and the CAP advice on teeth whitening, which covers evidence, lasers and who may carry out the treatment, does not use the word enamel²⁴²³.

"ADA" on a product usually points to the American Dental Association's Seal of Acceptance. The association describes it as a mark its Council on Scientific Affairs awards to over-the-counter products whose makers submit laboratory or clinical data against its requirements; it is an American scheme, and the hydrogen peroxide entry of the UK regulation makes no reference to it⁷. The nearest UK counterpart is the Oral Health Foundation's product accreditation, which says an independent expert panel checks that a product's claims are backed by reliable scientific evidence²⁶. In September 2026 its approved-products page linked accreditation pages for four brands, and none of them names a whitening or bleaching product; the foundation says more than 150 approved products are on sale, this page could not open a full list, and so it cannot rule out an accredited whitening product elsewhere²⁶.

Anyone looking for fluoride in a whitening product should know that the trials that tested fluoride around bleaching used professional gels¹³¹². In a double-blind randomised trial of 30 people bleaching at home with 16% carbamide peroxide, a 1.23% sodium fluoride gel worn in the tray for four minutes after each session did not change whether sensitivity happened, made it less intense and left the colour change intact¹³. In a triple-blind split-mouth randomised trial of 32 patients, 29 of whom finished, adding 2% sodium fluoride to a 40% in-chair gel lowered only sensitivity to cold, and only at 90 days, and its authors concluded that it did not demonstrate a benefit for bleaching sensitivity¹². Neither was a toothpaste, and neither result carries over to the fluoride in an everyday paste.

Where does S3 sit among the toothpastes people use around a course?

Beside the other daily sensitivity toothpastes, as a paste with declared actives and full-strength fluoride that no trial has put into either technique.

The toothpastes tested around a course have a patchy record. Thirty-eight adults brushed with a 5% potassium nitrate toothpaste or a placebo paste for a week before in-chair bleaching and before each session, in a triple-blind randomised trial, and the groups did not differ in sensitivity or in colour²⁷. An arginine sensitivity toothpaste worn in a tray for four minutes a day, from a week before in-chair bleaching with a 38% gel, made no difference to sensitivity or shade in a double-blind randomised trial of 48 patients²⁸. Across five trials and 387 people, a 2024 systematic review found that the answer turned on the protocol: desensitising pastes helped alongside a single 35% in-chair session and alongside 22% carbamide peroxide at home, and did not help alongside two in-chair sessions or 16% carbamide peroxide at home²⁹. The evidence on potassium nitrate before a course is set out on the page on potassium nitrate before bleaching, and using a sensitivity toothpaste during one on the page on whitening with a sensitivity toothpaste.

S3's product page declares 5% potassium nitrate beside two hydroxyapatites, nano at 10% and biomimetic at 5%, both inclusion levels of a solution, and sodium monofluorophosphate giving 1450 ppm fluoride. Its published ingredient list contains no hydrogen peroxide, no carbamide peroxide and no other peroxide. In everyday dentine sensitivity the three are meant to split the work, with potassium acting on the nerve, the nano form of hydroxyapatite inside the tubule and the biomimetic form on the surface. Sensitivity during bleaching is thought to come about differently, as the page on why whitening makes teeth sensitive explains, so a result in one is not a result in the other. No toothpaste, S3 included, has been tested as a way of making in-chair bleaching or home trays hurt less, and S3 has not been tested alongside either.

The table lists what each kind of paste declares and what the bleaching trials found, with S3 as a row taken from its own published ingredient list.

What should you ask your dentist?

Four questions cover the things the evidence says change the pain, and each one points back to a row of the results table.

1. What strength of gel will you use? The trial of 6% against 35% and the 2020 concentration review are the reason to ask: lower strength meant less sensitivity, and in the trial it also meant less colour⁴².
2. How long will the gel stay on, and can we shorten it if it hurts? Shorter home wear cut sensitivity events in the 2024 meta-analysis, at some cost in colour, which is a trade worth agreeing before you start³.
3. Is a light part of the treatment, and what is it for? A meta-analysis and an umbrella review found no gain in whiteness or comfort from light or laser activation; if the light is meant to ease pain afterwards, that is photobiomodulation, which is a different thing¹⁷⁵.
4. My teeth are already sensitive: do you know why, and does that change the plan? In the practice-based study, sensitivity present at the start predicted side effects, and the 6% trial left people with a history of sensitivity out altogether¹⁹⁴. The NHS and the Oral Health Foundation both put a check of the teeth and gums before any whitening⁶¹⁰.

Frequently asked questions

Is laser whitening worse for sensitive teeth?

Not clearly worse, and not gentler either. A meta-analysis of 21 randomised trials found that a light made no difference to sensitivity, and the 2026 umbrella review found activation less favourable with some high-energy protocols¹⁷⁵. The gel under the light matters more: in the double-blind trial of 140 adults, 74% had sensitivity on 35% and 44% on 6%⁴.

How long does sensitivity last after in-chair whitening compared with trays?

Usually days, and where the chair made a difference it showed early. In the triple-blind trial run with manufacturer-donated gels, chair patients hurt more only on the first day, and in the combined-course trial the extra sensitivity from the chair session sat in the first week²⁰⁸. The NHS says side effects may last a short time and the Oral Health Foundation calls them usually short-term⁶¹⁰. At a four-and-a-half-year follow-up of an in-chair trial with 35% gel, 25.5% of 62 patients reported long-term sensitivity, but nobody in it went unbleached, so that figure cannot be put down to the bleaching³⁰.

What does "enamel-safe" mean on a whitening product?

It has no UK definition that this page could find, and the ASA website's search turned up nothing under the phrase in September 2026²⁴. In laboratory studies, professional peroxide bleaching reduced enamel hardness slightly, with no clear evidence that the change matters clinically²⁵.

What does S3 contain, and is any of it a bleaching agent?

S3's declared actives are potassium nitrate and two forms of hydroxyapatite, nano and biomimetic, with 1450 ppm fluoride as sodium monofluorophosphate. None of it is a bleaching agent: its published ingredient list contains no peroxide of any kind.

Is the ADA seal relevant in the UK?

It is an American Dental Association scheme, and the hydrogen peroxide entry of the UK regulation does not mention it⁷. The UK's own product accreditation is run by the Oral Health Foundation, and the accreditation pages this page could open named no whitening or bleaching product²⁶.