Sensitive teeth not improving? A four-week troubleshooting plan that ends in a decision, not a verdict.

A four-week troubleshooting plan cannot speed up a sensitivity toothpaste, but it can rule out the pain patterns that need a dentist, fix the routine and the record, and find the mechanism your tube may be missing, so that week four ends in a clear decision instead of another guess. The verdict on any paste still comes at eight weeks, the length the consensus guidelines for sensitivity trials set for most studies¹, and an Italian expert consensus found that nobody had ever defined when each treatment step should happen². S3 Sensitivity Science™ asks to be judged at two to four weeks, which is sooner than the eight the trials use, so this plan stops at four weeks in a decision and not in a verdict on S3 or on any other tube.

What was checked24 peer-reviewed studies, reviews and consensus reports, guidance from the NHS, the Oral Health Foundation and GOV.UK, and S3's own published wording

What does the plan look like, week by week?

Five dated steps, a decision at week four and a verdict at week eight, each with the evidence behind it written next to it. Where a trial or a systematic review exists, the table says so; where a health body has written advice, it says guidance; where specialists agree without a trial, it says expert consensus; and where nothing but the logic of the step stands behind it, it says none. Several rows read guidance or none, and they are meant to.

The week-three row names S3 as one example of a paste that carries a nerve active and hydroxyapatite in the same tube. Every other row would read the same whatever you brush with.

Before you start: which signs mean a dentist now, not in four weeks?

Any of the signs in the table below, on any day of the plan. A plan built around a paste assumes the pain is the short, sharp twinge of exposed dentine; each row describes something else, or something that has gone on long enough for the guidance to want an examination whatever the tube is doing.

One row needs a second look before day one. The Oral Health Foundation tells people to see a dentist when sensitivity lasts more than a few weeks⁷, and someone reading a page about a paste that is not helping has usually crossed that line already. On that reading the plan runs beside an appointment, never in place of one, and day one is a sensible day to book it.

The table is arranged by what you notice, what the guidance says to do and how soon. The last column points to the page on this site that explains each pattern; the Journal's article on why teeth can turn sensitive suddenly covers the sudden kind.

Day one: what should you write down, and what does your tube actually contain?

Two things: a baseline to hold week four against, and the actives printed on the tube. Neither takes long, and together they stop week four turning into an argument with your memory.

The record is borrowed from the trials, and the page on how long to give a new sensitivity toothpaste sets it out in full: two triggers you meet anyway, such as a cold drink from the fridge and a breath of cold air, one scale, and each score written down on the day. This plan only fixes the dates. Day one is the first entry. Weeks two and four come next, and week eight belongs to whichever paste you finish on.

Why write it down at all? Because a tooth held to the same trigger is steadier than the days around it. Twenty-nine adults were tested again and again over three weeks in one clinical study, and the cold-air temperature that made a sensitive tooth hurt moderately to strongly hardly moved between sessions for the same person, even though it differed a lot from person to person⁸. The days are another matter. A fortnight's diary tells the other half: in a study funded by GlaxoSmithKline Consumer Healthcare, observational and not a trial, 101 self-diagnosed people at a UK university logged their twinges every evening, and one person's count could rise and fall sharply between days, with lower mood tracking worse pain¹⁷. A record read against its first entry averages out that noise; a verdict reached on a bad Thursday does not.

One rule comes straight from the trial method: do not test the same tooth twice in a row. In a clinical study of 80 volunteers, a cold air-blast needed about two minutes before a tooth gave the same answer again¹⁸.

Then read the tube. Calm the nerve but leave the tubules open and the triggers keep arriving; seal the tubules but leave the nerve over-reactive and it still complains. The ingredient panel tells you which of the two your tube is working on, and that is what week three will look for. Copy the active or actives, the percentage if the pack prints one, and the fluoride content in parts per million. The page on finding an equivalent by active ingredient shows which printed fields to copy, and the category review of UK sensitivity toothpastes by ingredient lists what each brand declares. For the fluoride line, the NHS asks adults to use a toothpaste with at least 1,350 parts per million⁹.

Last, note anything that changed in the month before the sensitivity rose: a whitening product, a new diet, a filling, a scale and polish, a new brush. Those are the causes on the page on what to rule out when sensitivity keeps climbing, and this plan does not rebuild that page. It gives each cause a day.

Week one: which routine changes have guidance behind them, and which are folklore?

Four have guidance behind them. Three of those rest on keeping fluoride in the mouth to prevent decay, and the fourth, an hour between acid and the brush, is advice for sensitive teeth whose evidence is about wear in the laboratory, not pain. None of the four has been tested as a treatment for sensitive teeth, which is worth knowing before you expect one week of them to show up in the record.

Brush twice a day with a fluoride toothpaste, last thing before bed and at one other time, for about two minutes⁹. Behind that advice sits a measured effect on decay: Delivering Better Oral Health, the prevention toolkit for dental teams published on GOV.UK, says that moving from brushing once a day to twice lowers a person's risk of decay by 14%¹⁹. After brushing, spit out the excess and do not rinse with water, because rinsing washes away the concentrated fluoride the toothpaste leaves behind⁹. The toolkit is candid about how strong that part is: higher-fluoride paste, twice-daily brushing and not rinsing, taken together, "appears to be associated with" less new decay in adolescents¹⁹. If you use a mouthwash, use it at a different time of day from brushing, such as after lunch, and the NHS applies that even to a fluoride mouthwash⁹²⁰. And leave at least an hour after acidic food or drink before brushing, which the Oral Health Foundation lists among its ways to reduce sensitivity⁷.

That hour is the one step in week one with a contradiction attached. A systematic review and meta-analysis of laboratory and in situ studies found no significant difference in the erosive wear of human enamel between brushing straight after an acid attack and waiting, and none of the studies it pooled had used human dentine, where sensitivity starts¹⁰. On the other side, in a cross-sectional clinical study of 350 adults in south-east England, sensitivity on the biting surfaces was more severe the more recently people had eaten or drunk something acidic, an association that cannot show the drink caused it²¹. So the hour is guidance with a plausible reason and mixed support. Keep it, and do not expect it to move the record on its own. The page on acidic food and drink triggers goes further into the acid itself.

On the brush, the NHS says medium or soft bristles suit most people, and that electric and manual brushes are equally good as long as every surface of every tooth gets cleaned⁹. If you suspect you press hard, that is one of the causes to rule out, and the page on what to rule out, linked above, weighs the evidence on force and abrasivity.

The folklore is most of what people try next: rinsing less, brushing for longer, using more paste, swapping the order of mouthwash and brush. For most of those, nobody has measured the effect on sensitivity at all, and the page on rinsing, mouthwash and timing after brushing sets out the little that has been measured, which is mostly fluoride and plaque.

Week one also sets the rule for the rest of the plan: change the routine, and leave the paste alone. The consensus guidelines for sensitivity trials ask for a negative control and a benchmark alongside the product under test so that whatever moves in everyone is not credited to the active¹. A person cannot run a control group. They can refuse to change two things in the same week, which at least tells them which change a week belongs to. If the paste changes in week one, an easier week two could belong to the routine, to the paste or to neither. Keep using it morning and night without gaps as well, because a missed week in the middle turns weeks two to four into noise. What stopping and restarting does to a sensitivity toothpaste is answered on its own page.

If the few-weeks row in the signs table applies to you, week one is when the appointment goes in the diary. A UK guideline review for general dental practice calls a joint working relationship between patient and dental professional, aimed at changing the patient's behaviour, essential if sensitivity is to be treated successfully²². Some of what week one asks of you is exactly that behaviour, and a dental team can see what a mirror cannot.

Week two: what should the first reading tell you, and what should it not?

It should tell you whether the record is being kept the same way, and whether anything from the signs table has appeared. It should not tell you whether the paste works, because at two weeks the trials are still watching every active climb.

In a double-blind randomised trial of 120 adults, funded by Procter & Gamble and run mostly by its employees, a potassium nitrate paste was ahead of a plain fluoride control on the cold-air score at week two, and further ahead at weeks four and eight¹¹. At week two, three in every hundred people on that paste had a test tooth that had stopped responding altogether; by week eight it was 28 in every hundred, in the same manufacturer-funded trial¹¹. A systematic review of 74 randomised trials, which grouped treatments by when their effect became significant, put potassium nitrate in its long-term group, and stannous fluoride and hydroxyapatite among those that got there by one month¹². Potassium-containing toothpastes are the most widely used home treatment, according to a 2006 review for practitioners²³, so on a potassium paste a small change at two weeks is what the evidence would lead you to expect, and not a sign that it has failed. Why that kind of relief builds instead of arriving is the subject of the page on relief that builds over weeks.

Two things can make week two feel different from what the tooth is doing, and both are explained on the page about whether it worked or only felt like it. Anyone who starts a paste in a bad week tends to find the following week easier whatever is in the tube, which is regression to the mean. People also shift the standard they judge their own symptoms by: in a secondary analysis of an eight-week trial of 75 adults, 36% moved theirs downward and 14.7% upward²⁴. That is why week two is compared with day one on paper, and why the plan never asks how the fortnight felt.

If the week-two entry reads clearly above day one on both triggers, skip to the last section of this page before going on.

Week three: which mechanism is missing, and where does a paste like S3 fit?

Read the actives you copied on day one and ask which job each does. Hydroxyapatite occludes and potassium desensitises, and neither does the other's job; potassium salts are there for the nerve, and most other desensitising agents work by occluding open tubules²⁵. If the label lists only a potassium salt, the tubule half is missing. If it lists only an occluding agent, the nerve half is missing. The page comparing tubule occlusion and nerve desensitisation sets out what each approach can and cannot do, and the Journal's article on how sensitive toothpastes work and why yours might not be working is the shorter read. For the two commonest single-active cases there are pages of their own: the page on potassium nitrate that is not working and six checks for a hydroxyapatite toothpaste that is not working.

The trials of adding the missing half are few and modest, and every one of them put both actives in a single formula, never in two tubes. In an industry-funded double-blind randomised trial of 127 adults, whose authors are employees of Lion, the maker of the pastes, or funded by it, adding aluminium lactate to a 5% potassium nitrate paste reduced sensitivity more than the potassium paste alone at four and eight weeks¹³. In a UK examiner-blind randomised trial whose record carries no funding statement, a paste combining potassium nitrate with hydroxyapatite and aluminium lactate beat a potassium-nitrate-only paste at every reading over 14 days, though with two ingredients added at once nobody can say which did the work²⁶. In an industry-funded double-blind randomised trial of 85 adults, in which Sangi made the nano-hydroxyapatite toothpastes, adding 5% potassium nitrate to a 10% nano-hydroxyapatite paste eased cold sensitivity further at weeks two, four and six, and not at week eight¹⁴. And a network meta-analysis of 125 randomised trials put potassium combined with hydroxyapatite among its largest estimates against plain fluoride toothpaste, on just two trials and one hundred and forty patients²⁷. None of those trials tested potassium nitrate, hydroxyapatite and fluoride together against their parts, and nobody has tested two tubes used side by side, which is the question the page on using two sensitivity toothpastes together takes on.

S3 is one paste that carries both halves: potassium nitrate for the nerve, nano-hydroxyapatite inside the tubule and biomimetic hydroxyapatite on the surface, with 1450 ppm fluoride kept in. Switching to it in week three is a real change of mechanism for someone whose tube had only one of the two, and it restarts the clock instead of shortening it: the day you open the new tube is a new day one, and its eight weeks start there. No trial has followed that exact combination of actives through the readings above, in any brand. Why one tube with two mechanisms differs from two tubes with one each is argued on the page on single-active and multi-active sensitivity toothpastes.

When is switching the wrong move?

When the record is improving, when both halves are already in the tube, and when a row of the signs table applies. In all three the next step is the same paste or a dentist. That holds whatever you are using and whatever this site would sell you: swapping one two-mechanism tube for another, S3 included, adds a variable without adding a mechanism.

Week four: continue, change, or book an appointment?

Read the week-four entry against day one and take one of three branches. If the record is improving, keep the same paste to week eight. If it is flat on a fair record, meaning the same two triggers, the same scale and no missed weeks, change mechanism and start again from a new day one, or book a dentist, and booking is never the wrong half of that choice. If your paste already carries both mechanisms, the change branch has nowhere to go, and booking is the branch left. If it is climbing, book, and do not try to fix it with another tube.

The three branches are not this page's invention. The Indian Society of Periodontology's 2022 consensus, written by an expert panel, sorts people at a recall visit the same three ways: continue an agent that has helped but not finished, change to another agent or add one with a different mechanism or move to treatment in the surgery when the complaint persists, and stop the active agent once the sensitivity has gone⁴. It was written for practice in India and is not UK guidance. Nor does it put a date on the recall for someone using a paste alone: its management charts write "Review" after home use of a desensitising paste with no interval beside it, and the only four-to-six-week recall in the document follows a desensitising mouthwash, or a paste and a mouthwash used together⁴.

Booking is a legitimate branch at any point, and what is known about what happens next is observational. In a practice-based study of 1,862 patients diagnosed by 171 dentists in a US research network, a potassium nitrate toothpaste was the commonest recommendation, made for half of them alone or with something else, and fluoride varnish came next at 28%²⁸. In the cohort from the same network, about six in ten patients reported that their sensitivity improved over eight weeks of whatever their dentist chose, and with no control group nobody can say how much of that the treatment did³. Neither is UK practice, and neither set treatments against each other. Together they say that an appointment is an ordinary step, and that it often ends with a paste recommended alongside something no tube provides.

What a dentist can offer, rung by rung and with how long each lasts, is on the page on dental treatments when toothpaste is not the answer.

Why four weeks, and why does the verdict still wait until eight?

Because four weeks is long enough to carry out every step that has guidance or a reason behind it, and eight weeks is where the evidence reads a paste. Four weeks is where this plan and S3's own wording stop, not where the evidence judges a toothpaste: the trial consensus uses eight weeks¹.

Nothing in the published protocols chose four weeks for this page. Every one read for it puts home care first and escalation second, and none prints four weeks as the moment to decide about a toothpaste used on its own. The table sets out what each says about timing, and how much of each was read.

Two PubMed searches were run for this page on the tenth of September 2026, one for stepwise algorithms and protocols in reviews, guidelines and consensus statements, the other for review appointments, re-evaluation and follow-up visits measured in weeks. They returned six records and four, and the only papers among them that turned out to print a timing were the two consensus reports at the foot of the table. The one printed interval that would apply to someone using a paste on its own is the follow-up recommended by the advisory board whose meeting was funded by Haleon: one to three months, graded as good practice rather than drawn from a study⁵.

So where does four come from? From three things a reader can check. The Oral Health Foundation's line for sensitive teeth is more than a few weeks⁷, and its toothache page gives sharp pain from cold two weeks to improve before an appointment¹⁶; by week four both lines are behind you. Four weeks is halfway to the eight at which trials read a paste¹. And it is the far end of the window S3 gives for itself, which is exactly why this page will not let week four become a verdict on S3. The NHS's two days, the check-up intervals and the other clocks a UK reader meets are compared on the page on how long to try a sensitivity toothpaste before seeing a dentist; this plan uses only the ones above.

There is a last reason, and it costs this site something to print. Several steps in the plan deal with things no paste touches: S3's actives work on the nerve and the tubules, and none of them changes how long acid sits on a tooth, how hard a brush presses, or what a crack is doing inside one.

What if it is getting worse at any point?

Stop running the plan as a plan, and book a dentist. A record that reads clearly above its day-one entry on both triggers, two readings running, is information a fourth week will not change. Changing the paste at that point adds a new variable to a problem nobody has diagnosed yet.

Nobody has published a threshold for how large a rise has to be before it counts, so decide on day one what would send you to a dentist, and write it beside the first entry. A rising record is a different question from a flat one, and the page on what to rule out when sensitivity keeps climbing, linked from day one, works through the causes no tube reaches: acid, brushing force and abrasivity, a gum line that has moved, and pain that is not sensitivity at all.

Carry on brushing with the same paste while you wait for the appointment. Bring the written record along: the NHS describes a check-up as starting with questions about your general health and about anything that has troubled your teeth, mouth or gums since you were last seen³¹. What happens in the chair is covered by the guide to what a dentist checks for sensitive teeth, and the cost by the page on free and low-cost NHS dental options. And if any row of the signs table appears along the way, it overrides the plan on the day it appears.

Frequently asked questions

If I switch to S3 in week three, when should I decide whether it is working?

Not in week four. Switching restarts the record, so the day you open the new tube is a new day one, with the same two triggers and the same scale. S3's own wording asks to be judged at two to four weeks from that day, and the trials read any desensitising paste at eight¹. Treat week eight as the verdict, and the two-to-four-week mark as a first reading and a check that nothing from the signs table has appeared.

S3 comes with a money-back guarantee, on the terms published on the product page. A guarantee on published terms is a commercial arrangement, and it does not move the point at which the evidence reads a paste.

Should I change toothpaste in week one?

No. Week one changes the routine, and a paste changed in the same week leaves an easier week two with two possible owners and no way to tell them apart. The consensus guidelines for sensitivity trials ask for a negative control and a benchmark beside the test paste for the same reason¹. The paste changes in week three, if at all, and only when the label shows a mechanism missing. If a dentist has recommended a different paste, follow that advice and start the record again from that day.

What if my teeth feel worse in week two?

Compare the entry with day one before you trust the feeling. A single bad reading after a week of cold drinks, a stressful few days or a missed evening brush is the kind of swing that a diary study of 101 people, observational and not a trial, funded by GlaxoSmithKline Consumer Healthcare, recorded from one day to the next¹⁷; two readings running above day one on the same triggers are a trend. If the trend is up, or any sign from the signs table has appeared, book a dentist instead of waiting for week four.

Do I need a dentist if the plan works?

Yes, for the ordinary reason. The NHS says the time between check-ups can vary from three months to two years, depending on how healthy your teeth and gums are, and that problems between check-ups are a reason to contact your dental surgery³¹. A plan that settled your sensitivity has not examined the tooth, and the Oral Health Foundation notes that sensitive teeth can sometimes be a sign of a dental problem⁷.

How long should I keep going if it is slowly improving?

To week eight on the same paste. Eight weeks is where the trial consensus reads a paste¹, and in the industry-funded trial described above, paid for by Procter & Gamble, every active arm gained again between week four and week eight¹¹. How to score those later weeks is on the page on how long to give a new sensitivity toothpaste, linked from day one, and what comes after is on the page on managing tooth sensitivity long term.