Switching sensitivity toothpaste: how long to give the new one before judging it.
Give a new sensitivity toothpaste eight weeks before you decide it has not worked, because eight weeks is where the field's own consensus guideline tells researchers to take the verdict on one1. Two weeks cannot settle it: in a six-month randomised placebo-controlled crossover trial of 35 patients, water alone cut air-blast pain by 20% immediately and by 36% at six months2. S3 Sensitivity Science™ asks to be judged at two to four weeks, which is earlier than the trial guideline's eight.
What was checked20 peer-reviewed studies, S3 consumer trial (ADSL, 2026), product information as published by each brand, guidance from the Oral Health Foundation and the NHS
- The consensus guideline for trials in this condition asks for a tactile stimulus and an evaporative one, a negative control, a benchmark control and eight weeks; those eight weeks are the field's judging point, not a counsel of patience1.
- In an industry-funded double-blind trial that read the same two instruments four times, a potassium nitrate paste's advantage over a control paste on the cold-air score went 13% at day three, 20% at week two, 35% at week four and 44% at week eight; for scale, water alone in a randomised placebo-controlled crossover trial of a paste applied in the surgery cut air-blast pain by 20% immediately23.
- S3's own guarantee is written at two to four weeks, which is earlier than the trial guideline judges any desensitising toothpaste.
- Nobody has published a head-to-head of moving from one desensitising toothpaste to another, so every date on this page is borrowed from work on a single paste against a control.
- In a secondary analysis of an eight-week trial of 75 adults, 36% had shifted downward the internal standard they judged their own symptoms against and 14.7% upward, which is the argument for writing your score down rather than remembering it4.
How long should you give a new sensitivity toothpaste?
Eight weeks. The 1997 consensus guideline that trials in this field are still designed against asks for a tactile stimulus and a cold or evaporative one, a negative control and a benchmark control, most trials lasting eight weeks, and a follow-up afterwards to see whether the change holds1. Every item on that list exists to stop a researcher being fooled by a number that would have moved anyway. Only one of them can be borrowed at a bathroom sink, and it is the calendar.
There is a gap under all of this that nothing on the shelf will mention. Two PubMed searches run for this page in September 2026 — one for switching, crossover or a change of dentifrice, one for a washout period, a run-in or a previous dentifrice — returned fifty-nine and twenty-nine records between them, and none of them moved anybody from one desensitising paste to another. What exists instead is one paste at a time, given to volunteers and read against a control arm. So the schedule on this page is an inference: it describes what a new active did to a mouth that had not just come off a different active, which is not quite your situation.
The practical shape of the answer is a first reading at week two, a fair one at week four and the verdict at week eight, because that is where the instruments were read. Which active to move to is a different question, and the mechanisms behind it are set out in the Journal's page on how sensitive toothpastes work and why yours might not be working and, ingredient by ingredient, in the guide to switching. Settle that before the eight weeks start. This page is about the clock and the record.
What do the trials measure at day three, week two, week four and week eight?
Something at every one of them, and less at the early ones than you would hope. The cleanest source is an industry-funded double-blind trial of 120 adults, paid for by the maker of the stannous fluoride paste it tested, which read a cold-air score and a probe threshold at day three and again at weeks two, four and eight3. What follows is read off its tables rather than its abstract.
On the cold-air score, all three active pastes in that industry-funded trial were ahead of the fluoride control paste from day three, by 18% for stannous fluoride, 20% for an oxalate paste and 13% for potassium nitrate; the same three figures were 36%, 28% and 20% at week two, 48%, 38% and 35% at week four, and 57%, 47% and 44% at week eight3. By that last visit, in the maker's own trial, the three actives were no longer significantly apart from one another3. On the probe, which raises a measured force against the tooth until you feel it, potassium nitrate's day-three figure of 10.7% did not reach significance in that manufacturer-funded trial, and it separated from the control only from week two3. The maker's trial also counted how many people had at least one test tooth stop responding altogether: on potassium nitrate, nobody at day three, one person in thirty at week two, two at week four and eight at week eight3.
Other actives run on their own clocks. In an examiner-blind randomised trial funded by Haleon, which makes the paste, and written by its own scientists, a bioactive glass toothpaste beat a plain fluoride paste from day three and the gap widened progressively to day 565. In a double-blind randomised trial of 85 adults sponsored by the maker of the nano-hydroxyapatite pastes it tested, scores fell at every fortnightly reading to week eight6. And a systematic review of 74 randomised trials grouped every treatment by the window in which its effect first reached significance: the up-to-seven-days group holds only treatments a dentist applies, and potassium nitrate appears in the long-term group alone7. That active's own record, reading by reading, is laid out on the page on whether potassium nitrate works within a week. The newest network meta-analysis, of 93 randomised trials, builds its whole comparison at two weeks, which is as early as the trials share a common reading8.
The last row of the table below is ours, and it is the thinnest one on it.
| Time point | What was measured there | What it does not mean |
|---|---|---|
| Day three | Cold air already ahead of the control paste in the industry-funded double-blind trial: 18% for stannous fluoride, 20% for an oxalate paste, 13% for potassium nitrate. On the probe, potassium nitrate's 10.7% was not significant3 | That an active has started working for you. These are group averages against a control arm, and the smallest of them is a change one mouth would not reliably notice |
| Week two | Cold air 36%, 28% and 20% ahead of the control paste, in the same order, in the maker's trial. One person in thirty had a tooth stop responding3 | A verdict. Every arm of that manufacturer-funded trial was still climbing, and the newest network meta-analysis of 93 trials stops here8 |
| Week four | Cold air 48%, 38% and 35% in the same order, in the maker's trial. Two people in thirty had a tooth stop responding3 | That the curve has finished. In that manufacturer-funded trial every arm gained again between this visit and the next |
| Week eight | Cold air 57%, 47% and 44% in the same order, in the maker's trial, and the three arms no longer significantly apart. A tooth had stopped responding for 53%, 35% and 28% of users, against nobody on the control3 | That a gap you can read across a row is a gap you would feel. At this visit the trial's own statistics put the three actives level |
| S3's own actives | 5% potassium nitrate is the dose the placebo-controlled trials used. A network meta-analysis of 125 randomised trials puts potassium with hydroxyapatite among its largest estimates, on two contributing trials and one hundred and forty patients9. S3's own eight-week figure is a consumer one: 86% said they no longer avoid certain foods or drinks, up from 59% at 24 hours | A time course for the finished tube. No trial has followed this combination of actives through the four visits above, and the eight-week figure is what panellists said about themselves rather than a clinical measurement |
Why can a fortnight's improvement not settle it?
Because the response to being treated at all is large here, and it is largest early. In the six-month randomised placebo-controlled crossover trial of 35 patients above — of a paste applied in the surgery, not a toothpaste — the water arm cut air-blast pain by 20% immediately and by 36% at six months2. In a split-mouth subject-blind randomised study of 22 adults, a periodontal dressing laid over the sensitive teeth cut reported pain by 95% to a thermal stimulus and 85% to an evaporative one, and the authors concluded that the perception of pain here is altered by sensory factors10. A network meta-analysis of 30 randomised trials found a significant placebo effect running through this whole literature11.
None of which means you are imagining it. The improvement is real and it is felt; in two of those three, the people reporting it had genuinely been treated with something. It is simply not information about which tube you bought. The plainest demonstration is the oldest: in a six-week double-blind trial of 120 adults, potassium nitrate, strontium acetate and a plain fluoride paste all reduced sensitivity with no significant difference between them, and the plain fluoride group improved significantly from its own baseline on cold air12. Somebody in that trial brushed with a paste that had no desensitising active in it and got better12.
The consequence for your eight weeks is narrow and useful. Judge the eighth week against the first, not against the best day in between, and expect the opening fortnight to flatter whatever you have just started. Why the useful part arrives late rather than early is argued on the page about why relief that builds over weeks is a feature.
How would you judge S3, or any new tube, fairly?
By writing four numbers down, on four fixed dates, for two triggers you meet anyway. That is the whole method, and every part of it comes from something somebody measured.
Pick two triggers and keep them. A cold drink from the fridge is one; a breath of cold air across the front teeth, or the brush at the gumline, is the other. Do not provoke the same tooth twice in a row: in a clinical study of 80 volunteers, a tooth needed about two minutes to give the same answer to a second cold air-blast, and a touch stimulus needed longer still — the mean change in reported pain between an immediate repeat and a two-minute one was 8.0 for touch against 0.8 for cold air13. The same study found that scores vary far more between people than within one person, which is why your own first reading is the only comparison worth making13.
Score it the way the trials do. A scoping review of 71 studies found the visual analogue scale — a line from no pain to the worst you can imagine, marked with a cross — and the Schiff cold-air scale to be the two standard measures, often used together, and reported that the choice between them is justified inconsistently from paper to paper14. Either will do for one person. What matters is using the same one every time.
Then write it down, because your memory of week one will have moved by week eight. In a qualitative interview study of eighteen Australian adults with confirmed sensitivity, part of a research programme funded by a toothpaste manufacturer, seventeen had changed what they ate or how they ate it and about half no longer described those changes as restrictions at all15. It has been counted, too: in a secondary analysis of an eight-week trial, more than a third of participants had moved their own yardstick downward over the study and about one in seven had moved it up4. A yardstick that moves while you are using it is not a yardstick. A note written on the day is.
| Date | Trigger one score | Trigger two score | Anything else that changed |
|---|---|---|---|
| Day one, before the first brush | new tube, and which active it carries | ||
| Week two | |||
| Week four | |||
| Week eight |
Change one thing and the record means something; change three and it means nothing. S3 Sensitivity Science™ is a daily toothpaste rather than a fortnight's course of treatment, which is what its two-to-four-week judging point assumes, and it carries 5% potassium nitrate, the dose the placebo-controlled trials used. Its own eight-week figures come from an independent third-party consumer trial of 51 adults over eight weeks, a record of what panellists said rather than a clinical measurement.
The market leader gives the same advice about the clock from the other side of the shelf. Sensodyne states that it is important to continue brushing with its toothpaste twice a day, every day, for lasting tooth sensitivity relief, and that if pain or sensitivity still persists you should visit your dentist, as your sensitive teeth may indicate a different problem (category review, read in September 2026). Two brands with different formulas agree on the unit of measurement: months of twice-daily brushing.
This page compares ingredients and stated actions only, not clinical performance, based on what each brand states about its own formula. Prices and formulations may change; always check the pack.
What if it keeps improving after eight weeks — or stops?
It may well keep improving, and the little published past eight weeks is worth knowing before you close the book. A 2025 systematic review and meta-analysis of 22 randomised trials with at least six months of follow-up found significant pooled reductions for several desensitising agents, with glutaraldehyde and low-level laser giving the largest effects — and calcium-phosphate-based agents not reaching significance, which is the finding in that review least convenient for a page like this one16. In a 24-week single-arm trial of 75 adults funded by GSK Consumer Healthcare, with three of its five authors company employees, the examiner's cold-air score improved from week four while the quality-of-life questionnaire did not reach significance until week eight, and then kept falling at every four-week reading to week 2417. In an examiner-blind randomised trial of 76 adults, also funded by GSK and first-authored by one of its employees, two eight-week courses of a bioactive glass paste separated by eight weeks off gave the same small improvement as brushing with it without a break for the whole 24 weeks18.
Read together, those say something modest and useful. The examiner's instrument moves before the part you would call feeling better, so week eight is a floor rather than a ceiling; and the effect does not fall off a cliff the day the schedule ends.
One thing nobody can give you is the number you most want: how much improvement counts as enough. The search behind this page found no published threshold for the cold-air score or for a pain scale. The only one that exists belongs to a 48-item quality-of-life questionnaire, where a validation study pooling 311 participants from three earlier trials, funded by GlaxoSmithKline Consumer Healthcare and co-authored by two of its scientists, put the smallest important difference somewhere between 22 and 39 points19. Applied to a real eight-week trial, that threshold left 50.7% of participants improved on the clinical measure and 22.7% improved enough on the questionnaire for it to count4. So the threshold on this page is yours, not ours. Decide before you start what would make the tube worth buying again, write it beside the first score, and hold yourself to it in week eight.
When should you stop waiting and see a dentist?
Before the eight weeks are up, if the pain is not the kind a toothpaste is for. The Oral Health Foundation, on a page carrying an educational grant from a toothbrush manufacturer, lists four reasons to book: pain that is severe, sensitivity lasting more than a few weeks, one tooth only, or pain that arrived suddenly — any of which can point to decay, a cracked tooth, gum problems or infection20. The NHS is shorter still: toothache lasting more than two days is a question for a dentist, as is toothache that painkillers do not touch21.
Those thresholds and this page's eight weeks are not in conflict, because they describe different symptoms. Eight weeks is the judging window for the short, cold-triggered twinge you have had for a while and can predict. Anything new, sharp, one-sided or lingering is not on a schedule at all. It is worth being honest about how much of the day this occupies: in a cross-sectional study of 3,551 adults across seven European countries, funded by Haleon, 29.1% had at least one tooth scoring moderate or severe to a cold-air stimulus, and just over half of those with the condition rated the pain as important or very important to their lived experience22. Waiting eight weeks is a reasonable thing to ask of someone. Waiting eight weeks on the wrong problem is not.
Frequently asked questions
Can I judge a sensitivity toothpaste after one week?
No, and the literature is unusually clear about why. The systematic review of 74 randomised trials puts nothing sold in a tube in its up-to-seven-days window; the only treatments that reached significance there are applied by a dentist7. At the earliest reading any toothpaste trial takes, day three, the gaps from a control paste are real but small, and in that industry-funded trial one of them was not significant at all3. A week tells you whether you can live with the taste.
S3 offers a guarantee at two to four weeks — should I judge it at four weeks?
You can use both windows as long as you do not confuse them. The guarantee is a commercial arrangement on published terms, and four weeks is a fair moment to decide whether to keep paying for something. It is not the moment the evidence would pick: in the industry-funded trial that read four visits, every arm gained again between week four and week eight3. Judging at four weeks is judging early, on our own product as much as on anybody else's.
Should I use the new paste and my old one on alternate days?
No. It makes the record unreadable — a flat week could belong to either tube, or to neither — and nobody has published a test of alternating two desensitising pastes, so nothing can tell you what it does. Every trial behind the figures on this page gave one paste, morning and evening, and changed nothing else about the routine.
How should I score it at home?
Pick one scale and one pair of triggers, and keep both for the whole eight weeks. A line from no pain to the worst you can imagine, marked and dated, is what most of the published work uses14. Leave at least two minutes between testing the same tooth twice, and longer if the trigger is touch rather than cold, because the tooth needs that long to give the same answer again13.
What if it worked for two weeks and then slipped back?
Check the calendar before you blame the tube. Missed days and restarts are the commonest explanation, and what a break actually costs is answered on its own page. If the routine has been steady, look at the mouth before the tube: fresh recession, a new acid habit, a cracked filling. And a fortnight of improvement that faded may never have been the active at all — the placebo arms in this literature move that far on their own211.
Where S3 sits
The judging window S3 publishes for itself is two to four weeks; the trial guideline's is eight, and S3 is not exempt from the longer one, so a reader who judges any tube at a fortnight is judging it earlier than the literature would. It is sold as the toothpaste you use every day rather than a fortnight's course bought for a bad week. There is a money-back guarantee behind it, worded at two to four weeks, on the terms published with the product.
S3 Sensitivity Science™ pairs potassium nitrate with two hydroxyapatites and adult-strength fluoride in one daily paste.
See the toothpasteS3 Sensitivity Science™ carries a nerve-calming active, 5% potassium nitrate, with two forms of hydroxyapatite — 10% nano-hydroxyapatite and 5% biomimetic hydroxyapatite, both as solution — and full adult-strength fluoride, in one daily toothpaste. Calm the nerve, strengthen the enamel surface, protect against further wear: three actions in one tube. The formula is patent-pending S3 Repair Technology™, UK application GB2604755.5. Others are dentist recommended; S3 is dentist owned, with more than 20 UK dentists having invested their own money in it. Read more about S3.