Toothpaste for Sensitive Teeth: Which Active Works, and How

No one toothpaste is the answer for sensitive teeth. What the shelf offers is a handful of active ingredients that work in one of two ways, which a 2022 review of 138 trials sums up as calming the nerve inside the tooth or plugging the microscopic channels that carry cold and touch to it. Trials of most of them found some benefit over weeks, not days; the reviews that pooled those trials could not always tell the actives apart; and people given a paste with no active in it improved too. This page sets out which active works which way, what the trials measured, who paid for them, and how to use a paste so it has a fair chance.

Before choosing a paste

Ordinary sensitivity is a short, sharp pain when cold, sweet or a toothbrush reaches dentine, the layer under the enamel, after something has uncovered it — most often receding gums, acid wear or hard brushing. Why that happens, and how to tell it from other kinds of tooth pain, is on what causes sensitive teeth.

When it is not a toothpaste problem

Pain that lingers after the cold has gone, pain in one tooth, pain when you bite, swelling, or pain that wakes you at night is not what these pastes are for. Those patterns can mean decay, a crack or an inflamed nerve, and they need a dentist, not a different tube. The US labels say so in their own words; Sensodyne’s reads “Sensitive teeth may indicate a serious problem that may need prompt care by a dentist.”

The two ways a sensitivity toothpaste works

Dentine is threaded with tubules, fluid-filled channels running from its surface in towards the pulp, where the nerve is. When the surface is exposed, cold air or a probe moves the fluid and the nerve reports the movement as pain. A sensitivity paste either makes the nerve answer less, or stops the fluid moving.

Fig. 1 — Route a: potassium salts act at the nerve end of the tubule, so the fluid still moves but the nerve answers less. Route b: the other actives sit over or in the tubule openings, so less movement reaches the nerve at all.

a. Calming the nerve: potassium salts

Potassium, usually as potassium nitrate, is the nerve route. The 2022 review puts it plainly: potassium “acts by diminishing the excitability of pulpal nerves”. Nothing is sealed, so the tubules stay open. The Cochrane review of potassium pastes adds a caution about the mechanism itself: it rests on animal studies of deep cavities, a model the authors say “cannot be compared” with a person brushing their teeth.

b. Plugging the tubules: the rest

The same review names arginine, calcium sodium phosphosilicate and stannous fluoride as actives that “occlude the dentin tubules”, and says strontium “has a dual function”, blocking tubules and acting on the nerve. On a label, arginine appears alongside calcium carbonate, and calcium sodium phosphosilicate is the bioactive glass sold as NovaMin. Hydroxyapatite is sold for sensitivity on the same principle; its most favourable review bases the mechanism on laboratory work, saying it “has been shown in vitro to block dentinal tubules”. Whether it does enough in a mouth is the question the trials below try to answer.

What the trials actually measured

Sensitivity is a feeling, so trials measure it by provoking it. The two standard tests are a puff of air from a dental syringe, scored on the Schiff scale, as in a 2013 split-mouth trial, and a probe pressed on the tooth with increasing force, such as the Yeaple probe, which the 2022 review describes; the score is the force that first hurts, in grams in a 2010 trial of 122 people. The 2019 network analysis of 30 trials compared results at two, four and eight weeks, and the Cochrane review looked at six to eight. Most people’s idea of “it worked” is how their teeth feel with a cold drink, which is close to the air test but not the same thing, and the Cochrane review found that people’s own assessment of their sensitivity did not change significantly even where the air and probe scores did.

The awkward finding is how much people improve without an active. In a 1997 trial, 112 people used a strontium paste, a potassium paste or an ordinary fluoride paste for six weeks. All three groups improved, with no significant difference between them, and the authors read the fluoride group’s gains as “a substantial placebo effect”. The 2019 network analysis also found “a significant placebo effect”, and no significant difference between plain fluoride paste and placebo. So a before-and-after number means little on its own; what counts is the gap over a comparison paste, and that gap is smaller than the before-and-after figure.

What the reviews found:

The largest, a 2020 network meta-analysis of 125 trials in 12,541 people, found that most formulations beat placebo or an ordinary fluoride paste. Calcium sodium phosphosilicate was the most beneficial for all three stimuli it examined, with high to moderate certainty; stannous fluoride helped touch and air; arginine helped air; potassium and strontium helped touch, with moderate certainty. Its authors later reported, in the 2022 review, that hydroxyapatite on its own came out with low or very low certainty.
A 2018 meta-analysis of 53 trials found every active paste did better than a paste without one, except strontium and amorphous calcium phosphate, and rated the evidence for the rest moderate quality at most.
The 2019 network analysis is where the actives blur together: “No significant difference was detected” among calcium sodium phosphosilicate, potassium and strontium pastes, and none between potassium and placebo. It ranked nano-hydroxyapatite first at two and four weeks and arginine first at eight.
The Cochrane review of potassium pastes, now twenty years old, pooled six studies and found an effect on air and touch at six to eight weeks but not on people’s own assessment. It concluded that “no clear evidence is available” for potassium.

Taken together: several actives beat a dummy paste in trials of a few weeks, the reviews disagree about which is ahead, and none shows one active ahead on every test. A review that ranks one first is ranking probabilities within its own set of trials, and the next review, with a different set, ranks differently.

Who paid for the trials

The research group behind the largest review also studied who funds sensitivity toothpaste trials. Their summary, in the 2022 review: “the industry funded 58% of them, the funding failed to affect the directionality of results”.

For the papers on this page, the trial funding ledger shows something less tidy. For most of them we could not read a funding statement at all: the full text is paywalled, or sits behind a bot check we do not get around, and the ledger records that rather than guessing. Where we could read one, the 2015 placebo meta-analysis declares that “No external funding was available for this study”; the Cochrane review lists its authors’ university as its source, and states that one author “was rewarded a prize from Zendium Inc., Denmark in 1997”; and the most favourable hydroxyapatite review declares that two of its three authors “are senior scientists and employees of Dr. Kurt Wolff GmbH & Co. KG in Germany”, the company behind the hydroxyapatite trials discussed in does hydroxyapatite toothpaste work. None of the arginine trials cited here had a readable funding statement; the one competing-interests statement we could read, in the 2013 split-mouth trial, declares none.

A manufacturer paying for a trial does not make the trial wrong; it is context the reader is entitled to see next to the finding. Each paper’s ledger row is linked in the table below.

The actives, side by side

One row per active, in alphabetical order; the order means nothing. “In the database” opens our toothpaste database filtered to the labels that name that active. On a phone, the table scrolls sideways.

Active, as printed on labels	How it works	How soon it may help	What the evidence shows	Who funded most trials	In the database
Arginine, with calcium carbonate; usually “8%”	Tubules	Relief straight after rubbing it onto the tooth, and after 3 days of brushing, in the 2010 trial; one trial, which began with a paste applied at the dentist, ran 24 weeks	Helped air-blast pain compared with ordinary fluoride, moderate certainty (2020 review). An earlier review saw “a potential role”, from small trials with no long follow-up	Not readable for the three trials cited (ledger: Fu 2010, Kapferer 2013, Hirsiger 2019)	6 labels
Calcium sodium phosphosilicate, sold as NovaMin	Tubules	Compared at 2, 4 and 8 weeks (2019 analysis)	The most beneficial for touch, cold and air in the largest review, high to moderate certainty; no different from potassium or strontium in the 2019 analysis	Reviews only here; their own funding not readable (ledger: Martins 2020, Hu 2019)	3 labels, NovaMin or BioMin; none among the US labels read
Hydroxyapatite, often “nano-hydroxyapatite”	Tubules (laboratory evidence)	An independent review pooled trials that followed people for 4 weeks	Helped air and touch but not cold over 4 weeks, with longer trials still needed; moderate-quality evidence in the 2018 review; low certainty on its own in the largest	The most favourable review is co-written by a manufacturer’s employees (ledger: Limeback 2023)	21 labels
Potassium nitrate; US labels state 5%	Nerve	Effect on air and touch at 6 to 8 weeks (Cochrane)	Helped touch pain, moderate certainty (2020 review); no significant difference from placebo in the 2019 analysis; Cochrane found “no clear evidence”	Cochrane review: the authors’ university; the 1997 trial not readable (ledger: Poulsen 2006, West 1997)	18 labels
Stannous fluoride; US labels state 0.454%	Tubules	The reviews we read give no separate time course for it	Helped touch and air pain, high to moderate certainty (2020 review); moderate-quality evidence in the 2018 review	Reviews only here; one declares no external funding (ledger: Bae 2015)	14 labels
Strontium chloride or acetate	Both (2022 review)	No better than ordinary fluoride paste at 2 or 6 weeks in the 1997 trial	Helped touch pain, moderate certainty (2020 review); no significant effect in four pooled trials in the 2015 meta-analysis	One trial cited, funding not readable (ledger: West 1997)	None of the 126 labels we read names a strontium salt

Got a tube at home?

Paste its ingredient list into the toothpaste ingredient decoder to see which sensitivity active it names, if any. And if a brand’s page says “clinically proven”, the reference checker sorts the studies it cites so you can see what kind of evidence stands behind the phrase.

How to use it so it has a chance of working

Twice a day, as your ordinary toothpaste. The US labels we read direct brushing “for at least 1 minute twice a day (morning and evening)”. Swapping between pastes means none of them gets the use it was tested under.
Spit, don’t rinse. England’s prevention toolkit advises “spitting out after brushing rather than rinsing with water, to avoid diluting the fluoride concentration”. Its reason is the fluoride; none of the trials we read tested rinsing against the sensitivity active itself.
Give it weeks, not days. The reviews compared results at two, four and eight weeks, and some US labels tell you to stop and ask a dentist if “pain/sensitivity still persists after 4 weeks of use”. Four days is not a trial.
Keep the fluoride. A few sensitivity pastes are fluoride-free. For adults the toolkit’s advice is a toothpaste “containing 1,350 to 1,500ppm fluoride”, and the database shows each label’s figure.
One sore spot. In the 2010 arginine trial, which measured the quickest relief, people put “a pea-size amount” on the sensitive surface and “massaged for 1 minute”. That is the method behind the fastest results. One tooth that is clearly worse than the others is also the pattern to show a dentist.

What to be wary of

“Instant relief”. The trials behind the phrase measured something narrower than it suggests. In the 2010 trial, people rubbed an 8% arginine paste onto the sore tooth with a fingertip for a minute; in the 2013 trial, the paste was applied in the dental chair and was not a toothpaste at all. Both found less pain straight afterwards than with a calcium carbonate paste. That is a targeted application measured with air and a probe, not a promise about brushing. One US carton in our database says “Sensitivity Relief starts in just 1 Day”; its label names stannous fluoride as the only active (Colgate Sensitive Pro-Relief Whitening).
“Repairs enamel”. The largest review measured pain from air, cold and touch. None of the reviews on this page measured enamel. A paste that reduces sensitivity in these trials has shown that, and only that.
High-abrasive and whitening pastes. For people with tooth wear, the toolkit says to “avoid using high abrasive toothpaste”. Most labels do not let you check: 120 of the 126 in our database state no abrasivity figure. Whitening pastes mostly work by abrasion, as natural teeth whitening explains.
The same name in another country. Sensodyne Repair and Protect names NovaMin on its UK page; Sensodyne Repair & Protect lists stannous fluoride on its US label. Sensodyne Pronamel Daily Protection discloses no active on its UK page; Sensodyne Pronamel Daily Protection lists potassium nitrate. A review of one is not a review of the other.
Stannous fluoride and staining. Every US stannous fluoride sensitivity label we read on DailyMed carries the same note: such products “may produce surface staining of the teeth”, which the label says brushing may prevent and “may be removed by your dentist” (Sensodyne Rapid Relief).
Two potassium salts on one label. Colgate Sensitive with Sensifoam Multi Protection lists potassium nitrate and tetrapotassium pyrophosphate. The nitrate is the one named for sensitivity; the pyrophosphate is there as a tartar-control salt.
A “sensitive” paste with none of these actives. The list for elmex Sensitive names no potassium salt, stannous fluoride, arginine, bioactive glass, hydroxyapatite or strontium. Its fluoride is sodium fluoride at 1450 ppm.

The first of these is now a card in the Marketing or evidence? quiz, if you want to test the instinct on other claims.

How we made this page

Every product mentioned is a row in the toothpaste database, read from the manufacturer’s own page or the label it filed with the US Food and Drug Administration, and dated; the database explains how each row is read. Every review and trial cited below has a row in the trial funding ledger, with its funding statement quoted rather than summarised. We did not test any product ourselves, and nothing here is personal advice: a dentist who has looked at your teeth can tell you things this page cannot.

We don’t link to shops from this page and earn nothing from it.

Fig. 1 is free to reuse under CC BY 4.0 — see the diagrams page.

References

Martins CC, Riva JJ, Firmino RT, et al. Formulations of desensitizing toothpastes for dentin hypersensitivity: a scoping review. J Appl Oral Sci 2022;30:e20210410. DOI · PubMed · Funding: see ledger
Haleon US Holdings LLC. Sensodyne (potassium nitrate and sodium fluoride) toothpaste, Drug Facts label, as filed with the US Food and Drug Administration. DailyMed. Official page, accessed 23 Sep 2026.
Poulsen S, Errboe M, Lescay Mevil Y, et al. Potassium containing toothpastes for dentine hypersensitivity. Cochrane Database Syst Rev 2006;(3):CD001476. DOI · PubMed · Funding: see ledger
Limeback H, Enax J, Meyer F. Clinical evidence of biomimetic hydroxyapatite in oral care products for reducing dentin hypersensitivity: an updated systematic review and meta-analysis. Biomimetics (Basel) 2023;8(1):23. DOI · PubMed · Funding: see ledger
Kapferer I, Pflug C, Kisielewsky I, et al. Instant dentin hypersensitivity relief of a single topical application of an in-office desensitizing paste containing 8% arginine and calcium carbonate: a split-mouth, randomized-controlled study. Acta Odontol Scand 2013;71(3–4):994–999. DOI · PubMed · Funding: see ledger
Fu Y, Li X, Que K, et al. Instant dentin hypersensitivity relief of a new desensitizing dentifrice containing 8.0% arginine, a high cleaning calcium carbonate system and 1450 ppm fluoride: a 3-day clinical study in Chengdu, China. Am J Dent 2010;23(Spec No A):20A–27A. PubMed · Funding: see ledger
Hu ML, Zheng G, Lin H, et al. Network meta-analysis on the effect of desensitizing toothpastes on dentine hypersensitivity. J Dent 2019;88:103170. DOI · PubMed · Funding: see ledger
West NX, Addy M, Jackson RJ, et al. Dentine hypersensitivity and the placebo response. A comparison of the effect of strontium acetate, potassium nitrate and fluoride toothpastes. J Clin Periodontol 1997;24(4):209–215. DOI · PubMed · Funding: see ledger
Martins CC, Firmino RT, Riva JJ, et al. Desensitizing toothpastes for dentin hypersensitivity: a network meta-analysis. J Dent Res 2020;99(5):514–522. DOI · PubMed · Funding: see ledger
Hu ML, Zheng G, Zhang YD, et al. Effect of desensitizing toothpastes on dentine hypersensitivity: a systematic review and meta-analysis. J Dent 2018;75:12–21. DOI · PubMed · Funding: see ledger
Bae JH, Kim YK, Myung SK. Desensitizing toothpaste versus placebo for dentin hypersensitivity: a systematic review and meta-analysis. J Clin Periodontol 2015;42(2):131–141. DOI · PubMed · Funding: see ledger
Hirsiger C, Schmidlin PR, Michaelis M, et al. Efficacy of 8% arginine on dentin hypersensitivity: a multicenter clinical trial in 273 patients over 24 weeks. J Dent 2019;83:1–6. DOI · PubMed · Funding: see ledger
Sharif MO, Iram S, Brunton PA. Effectiveness of arginine-containing toothpastes in treating dentine hypersensitivity: a systematic review. J Dent 2013;41(6):483–492. DOI · PubMed · Funding: see ledger
de Melo Alencar C, de Paula BLF, Guanipa Ortiz MI, et al. Clinical efficacy of nano-hydroxyapatite in dentin hypersensitivity: a systematic review and meta-analysis. J Dent 2019;82:11–21. DOI · PubMed · Funding: see ledger
Office for Health Improvement and Disparities. Delivering better oral health: an evidence-based toolkit for prevention (chapter 2, summary guidance tables, and chapter 7, tooth wear). Official page, accessed 23 Sep 2026. · Funding: see ledger

Toothpaste for sensitive teeth: what actually works, and why