What happens when you stop using a potassium nitrate toothpaste: the one trial that measured it.

When you stop using a potassium nitrate toothpaste, most of the relief does not go with it. One industry-funded double-blind randomised trial has deliberately stopped a potassium paste and measured what happened, and three weeks later the group that had been on it was still significantly ahead of its control¹ — a reading the placebo-controlled randomised trials behind the ingredient never took, because they finished with everyone still brushing⁶. That is not what the pack says, it is not what the guidance says, and it is not what this site said until today: the category, us included, has been repeating something nobody had checked. Of 51 UK sensitivity toothpastes whose ingredient lists were read for the category review in September 2026, S3 Sensitivity Science™ is the only one carrying potassium nitrate, hydroxyapatite and fluoride together.

What was checked11 peer-reviewed studies, product information as published by each brand, and NHS and Oral Health Foundation guidance; 5 PubMed searches

Does the relief actually go when you stop?

Less of it goes than everybody assumes, and one washout arm is the whole of what we know.

Start with who paid. Seven of the eight authors were employees of the company that funded the work, and the stannous fluoride paste on test is that company's own product¹. What that industry-funded group published in 2025 was an eight-week comparison of a marketed potassium nitrate toothpaste, that stannous paste and an experimental oxalate paste against a plain fluoride control — and then the unusual part: at week eight it took all four groups off their assigned tubes, put every one of them on the control paste for three weeks, and measured again at week eleven¹. The disclosure cuts an unexpected way here, and it is worth saying out loud: this particular finding is not the one a manufacturer would have designed the arm to produce.

At week eleven, in that manufacturer-funded trial, all three actives were still significantly ahead of the control group¹. Read the potassium nitrate column of the tables that manufacturer published and its improvement over the control on the cold-air score was 44.3% at week eight and 35.8% at week eleven, and on the touch threshold 104% at week eight and 72.5% at week eleven, with p below 0.001 at every one of those visits¹. Work the adjusted means against the control visit by visit and the potassium group kept roughly four fifths of its cold-air gain and roughly two thirds of its touch gain across three weeks with none of the active in its mouth; by week eleven the three actives were statistically indistinguishable from one another in the maker's own trial¹.

Then there is the shorter question, the one most readers are actually asking. A 2026 randomised comparison of 127 adults, every author of it an employee of or grant-funded by the manufacturer of the pastes on test, built a deliberate one-day gap into its schedule: after the four-week visit each participant brushed for a single day with the placebo paste and came back twenty-four hours later to be measured again, which its authors describe as a test of whether the effect persists after temporary discontinuation². In the arm using 5% potassium nitrate alone the manufacturer's tables record a touch threshold of 22.88 g at four weeks and 24.63 g a day later, and a cold-air score of 1.75 then 1.58 — both slightly improved after the missed day rather than worse². One caveat belongs in the same breath: the placebo group's readings in that manufacturer-run trial drifted the same way across the same twenty-four hours, so part of what looks like persistence is the measurement and not the paste².

A third trial points the same way with a different chemistry. In a double-blind randomised trial of 129 subjects, funded by a toothpaste manufacturer whose employees are four of its six authors, a paste combining hydroxyapatite with potassium citrate was used for eight weeks and then dropped for four, and the reduction in cold-air sensitivity was still there at the twelve-week reading³. Different salt, different mechanism: it corroborates the pattern, not the ingredient.

Against all of that sits the advice everyone has read. The Oral Health Foundation's page on sensitive teeth says of these toothpastes: "You need to keep using them to maintain the effect"⁴. Neither side of this is wrong. Guidance is written for a population nobody will measure, and telling that population to keep going is the sensible instruction; what was actually measured, once, turned out to be more forgiving than the instruction implies. The honest reading is that the advice is good and the single number we have is kinder than the advice suggests.

Can you get used to potassium nitrate?

Nobody knows, because nobody has looked, and that is the most useful sentence on this page.

Five searches were run against PubMed on the tenth of September 2026. The first repeated one from the day before — potassium, with dentine or dentin hypersensitivity, with tolerance, tachyphylaxis, loss of effect, relapse, long-term, withdrawal or washout — and returned fourteen records. A second swapped in desensitisation, desensitization, diminishing effect, reduced response and rebound: eleven records. A third asked for discontinuation, cessation or "after stopping" alongside the condition: ten records. A fourth asked for relapse, recurrence, return of symptoms or re-treatment: thirteen records. The fifth asked for tachyphylaxis, loss of efficacy or diminishing effect alongside toothpaste or dentifrice and sensitivity, and returned nothing at all — zero records. Forty-four unique papers came out of the four searches that found anything, and among them are lasers, varnishes, root-coverage grafts, bonding agents, a steroid-therapy case series and several papers on mouthrinses. Not one asked whether a potassium toothpaste stops working in the mouth of someone still using it.

That absence has a mundane cause. Potassium nitrate is measured over eight weeks, or twelve at the outside, and then the measuring ends: the 2006 Cochrane review pooled its six randomised trials at six to eight weeks⁵. A twelve-week placebo-controlled trial of a 5% paste is among the longest anyone has run⁶. Tolerance to anything is demonstrated by keeping people on it for far longer than that and watching the line come back down, and nobody in this field has bought that follow-up. So the word people reach for — "I built up a tolerance" — names something that has never been observed, tested or ruled out. What the eight-week data show is the opposite shape under continued use, with the potassium benefit larger at week eight than at week four in the manufacturer's own washout trial¹.

The empty rows are the most valuable cells in that table, and they are the reason this page exists.

So why did it feel as though it stopped?

Three explanations have evidence behind them, and none of them is the tube giving up.

The first is that the cause never stopped. A potassium paste acts on how readily the nerve fires and leaves the dentine tubule exactly as open as it found it. If the gum has carried on receding or the surface has carried on wearing, there is more exposed dentine this month than last, and a nerve being calmed to the same degree has more arriving at it. Nothing about the tube changed; the arithmetic on the other side of it did.

The second is the placebo component, and it deserves to be written carefully, because a placebo response is a real experience and not a delusion. In a six-week double-blind randomised trial of 120 adults, a potassium nitrate paste, a strontium acetate paste and a plain fluoride paste all reduced sensitivity, none of them separated from the others, and the plain fluoride group improved significantly from its own baseline on the cold-air test⁷. A 2019 network meta-analysis of thirty randomised trials measured a significant placebo effect across this field, and in its network potassium was not significantly separated from placebo⁸. The size of it has been put on the record directly: in a 56-day placebo-controlled randomised trial of 90 adults whose funding is not stated, the arm rinsing with a base formula stripped of its actives improved by 30% to 40%, which the authors present as a measurement of the placebo response in this condition rather than as a result about the rinse⁹. Thirty to forty per cent is a great deal of relief to feel, and then to lose. Expectation is at its highest when a new tube is opened and it wears off with familiarity, which produces exactly the story people bring to this question: it worked, and then it stopped.

The third is time, and it cuts both ways. A 2019 systematic review of 74 randomised trials grouped every desensitising treatment by the follow-up window in which its effect reached significance, and put potassium nitrate in the long-term group only; no toothpaste at all appeared in the seven-day window¹⁰. A tube given a fortnight was never given the timetable the evidence was built on. A separate page walks every published time point for this ingredient, from day three to week twelve, and the Journal covers the same ground from the other end in how sensitive toothpastes work.

Could the tube itself have changed?

Almost certainly not in a way you could check, and this page is not going to guess.

Based on what each brand states about its own formula, three products in the Sensodyne range whose ingredient lists were read at a UK retailer in September 2026 carry potassium nitrate. Pronamel Daily Protection declares potassium nitrate at 5% w/w alongside sodium fluoride at 1450 ppm; Daily Care Original and Complete Protection+ Original each list potassium nitrate among their ingredients and give no concentration for it. Across the whole scan the pattern repeats — most potassium nitrate products on the UK shelf publish the ingredient and withhold the level — so a reader who suspects a reformulation has nothing on the pack to compare.

What is comparable is the ingredient list itself. It is published, it is dated by the page it sits on, and the order of the entries carries information: hold the list on the tube in your bathroom against the one on the retailer page you bought it from, and a real change usually shows up as an ingredient that has moved, arrived or gone. Short of that, this page will not suggest that any product has been reformulated or diluted. That would be a factual statement about a company's manufacturing, it would need two dated ingredient lists that differ, and there are none.

Does a paste like S3, which also seals the tubule, hold its effect better?

Nobody has tested it, and it would be easy and dishonest to imply otherwise.

The mechanism argument is straightforward enough. Potassium desensitises the nerve; hydroxyapatite occludes the tubule; neither stands in for the other, and a tube built for sensitive teeth needs the pair. On that reasoning a paste that leaves mineral on and in the dentine surface is doing something physical that persists in a way an ion gradient around a nerve ending need not, and S3's own ingredient list carries 5% potassium nitrate with two forms of hydroxyapatite and full-strength fluoride.

Reasoning is not measurement. Nobody has taken a potassium-and-hydroxyapatite formula, stopped it, and measured what survived; the washout evidence that exists belongs to a potassium nitrate paste, a stannous paste, an oxalate paste and a hydroxyapatite–citrate paste, and the combination in S3 has never been through a published trial of any kind. The durability question is therefore open for S3 exactly as it is open for everything else on the shelf. Any page that tells you one paste's relief outlasts another's is describing something nobody has weighed.

When is "it stopped working" a dentist's question?

When the pain has changed character rather than simply faded. That is common enough behind this particular question to be worth checking before anything else.

A potassium nitrate paste was formulated for a short, sharp twinge that arrives with cold or sweetness and leaves when the stimulus does. If what you have now is an ache that outlives the drink, pain on biting, pain that wakes you, or pain in one tooth where it used to be several, the paste has not given up: the problem is no longer the one the paste was for. The Oral Health Foundation's threshold list runs to four items: severe pain, sensitivity that has run past a few weeks, one tooth rather than several, and an onset that was sudden. Behind them it puts decay, a crack, gum problems and infection⁴. The NHS sets its own bar at toothache of more than two days, or toothache that painkillers will not settle, or toothache arriving with a high temperature, pain on biting, red gums, a bad taste or a swollen cheek or jaw, and names an abscess, a crack, decay and a failing filling among the causes¹³.

None of those is a toothpaste problem. If any of them describes your month, book the appointment and take the tube with you; what you need next is a diagnosis, not a different brand.

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.

Frequently asked questions

Can you build up a tolerance to a sensitivity toothpaste?

There is no published evidence that you can, and none that you cannot, because the question has never been put to a trial⁵⁶. Five PubMed searches in September 2026 for tolerance, tachyphylaxis, habituation, loss of effect, diminishing effect, rebound and reduced response alongside potassium and dentine hypersensitivity returned forty-four unique records between them and not one on the subject; the search that asked for tachyphylaxis with a toothpaste returned nothing at all. Potassium nitrate is measured over eight to twelve weeks and then the measuring stops, which is too short a window to see tolerance even if it were there⁵⁶.

Will S3 stop working after a few weeks too?

No trial has tested tolerance to potassium nitrate in any product, S3 included, so the honest answer is that the question has not been asked of anything on the shelf. What has been measured, in the trials that ran to eight and twelve weeks, is a potassium nitrate effect that was still growing at the last reading rather than shrinking, and in the industry-funded trial with the washout arm it was still growing at week eight¹⁶.

If I stop for a fortnight, do I start again from the beginning?

Probably not, on the little evidence there is. The nearest measurement used a different active: an industry-funded examiner-blind randomised trial of 76 adults compared a bioactive-glass paste used continuously for 24 weeks with the same paste used in two eight-week courses separated by eight weeks on an ordinary fluoride paste, and the two schedules finished level on every outcome¹¹. Nobody has run that trial with potassium nitrate. Set beside the twenty-four-hour and three-week readings above, the evidence points to a slope rather than a cliff.

Why did it work for a month and then seem to stop?

The likeliest answers are a placebo component wearing off as novelty does, an underlying cause that carried on while the paste addressed the symptom, or a mouth that has changed since you started. A 2000 review of 27 clinical trials records that the proposed mechanism for potassium — ions gathering around the nerve endings and making them harder to set off — has never been confirmed in an intact human tooth, so nobody can tell you from first principles what a fading effect would even look like¹². What is worth ruling out first is the one item on that list a toothpaste cannot address: pain that has changed character belongs to a dentist⁴.