Probiotic oral lozenges: do they actually work?

Streptococcus salivarius K12 and M18 are the first oral probiotics with real clinical evidence. Here is what the trials actually show for bad breath, cavity-causing bacteria, and gum inflammation, plus how to dose them and what to ignore.

The pitch for oral probiotics is almost too tidy. The mouth, like the gut, has a microbiome of hundreds of bacterial species. Some are friendly, most are neutral, a small number drive cavities, gingivitis, and bad breath. Brushing, flossing, and mouthwash all try to knock down the bad ones, but they take out everything else in the same swing. Why not just add more of the good guys?

In principle, this is exactly what oral probiotic lozenges set out to do. In practice, the category is a mess. Most products on Amazon and in pharmacies are marketing exercises wrapped around generic strains with no published evidence behind them. A small fraction are based on two specific bacterial strains, isolated from healthy human carriers in New Zealand in the 1980s and 2000s, with several decades of clinical work behind them. This guide separates the two. It walks through what oral probiotics actually are, the BLIS K12 and M18 strain story, what the strongest trials show, where the evidence is genuinely thin, how to dose them properly, and how to read a label so you do not end up paying for a flavored sugar lozenge with a marketing claim attached.

What oral probiotics actually are

A probiotic, by the World Health Organization's working definition, is a "live microorganism that, when administered in adequate amounts, confers a health benefit on the host." That definition matters more than it looks. It implies three things: the organism has to be alive when it reaches the site of action, you have to take enough of it, and the benefit has to be documented in a host, not just on a petri dish.

For gut probiotics, this is reasonably well understood. The strains have evolved alongside the human intestine for thousands of generations, the dose ranges are mapped out, and the survival challenge is mostly about getting them past stomach acid alive. Oral probiotics are different in two important ways. First, the site of action is the mouth and throat, not the gut, so the strain has to be able to attach to oral surfaces (the tongue, the tonsils, the pellicle on tooth enamel) and to outcompete the resident flora locally. Second, the time the strain spends at the site of action is very short. A lozenge dissolves over 5 to 10 minutes; a swig of probiotic mouthwash takes 30 seconds. The window for the bacteria to find a home and start producing their useful compounds is small.

This is why the strain selection matters more than the genus. "Streptococcus salivarius" sounds generic, but most isolates of S. salivarius are passive members of the resident flora that simply exist alongside the others. The two strains that have made it into the oral probiotic literature, K12 and M18, are unusual within the species because they produce powerful bacteriocins. A bacteriocin is a protein toxin that one bacterium uses to kill or inhibit related species, usually those competing for the same niche. K12 produces salivaricin A2 and salivaricin B. M18 produces salivaricin M and salivaricin 9. These compounds give the strains a competitive edge against specific pathogens, and they are the reason the strains do anything at all.

It is also worth flagging what is not a probiotic, despite frequent claims to the contrary. Postbiotics (heat-killed bacterial fragments), prebiotics (substrates that feed existing flora, such as xylitol or certain fibers), and yeast products (like Saccharomyces boulardii in dental applications) are different categories with different evidence bases. They are sometimes useful, but they are not what the K12 and M18 trials are about, and conflating them is the single most common mistake in popular coverage of this category.

The K12 story: a child in New Zealand

Streptococcus salivarius K12 was isolated by John Tagg, a microbiologist at the University of Otago in Dunedin, New Zealand, in the late 1980s. Tagg had been working on streptococcal infections for years and had grown interested in why some children seem genuinely immune to strep throat while others get it repeatedly. He started screening saliva samples from healthy New Zealand children, looking for kids whose throats were colonized by strains that actively inhibited Streptococcus pyogenes (the bacterium responsible for strep throat) in lab assays.

The K12 strain was recovered from a child whose throat flora consistently killed S. pyogenes on culture plates. Genetic characterization showed that the strain carried two bacteriocin gene clusters (for salivaricin A2 and salivaricin B), encoded on a large megaplasmid that the strain could transfer to other susceptible S. salivarius cells. The strain was named K12 (Tagg has said the K is for "killer," in the sense of bacteriocin production). It was patented through the university's commercial arm and licensed to BLIS Technologies, a New Zealand company set up specifically to commercialize the bacteriocin-producing strains (BLIS stands for bacteriocin-like inhibitory substances).

From the early 2000s onward, BLIS K12 became the most studied oral probiotic strain in the world. The trials clustered around three applications. The first was prevention of streptococcal sore throat in children with recurrent infections. The second was reduction of volatile sulfur compounds in adults with halitosis (more on this below). The third was upper respiratory infection prevention, particularly recurrent acute otitis media (middle-ear infections) in children. Across all three application areas, the trials are small (typically 30 to 200 participants) but the results are reasonably consistent.

The key features of K12 that made it commercially viable: it is genuinely safe (no known clinical infections caused by the strain in over 20 years of monitoring), it survives the lozenge manufacturing process well, and it can transiently colonize the throat and tongue in a meaningful percentage of users. A 2006 paper from the Tagg group found that K12 could be detected in the saliva of about 50 percent of users at 28 days after a single 7-day course, although the colonization is rarely permanent without ongoing dosing.

What the halitosis evidence actually shows

Bad breath is the strongest indication for K12, in the sense that the clinical signal is the cleanest and the mechanism is the most plausible. Roughly 85 to 90 percent of chronic bad breath cases originate in the mouth itself, mostly from anaerobic bacteria living in the deep crypts of the posterior tongue and in periodontal pockets. These organisms (Solobacterium moorei, Atopobium parvulum, Eubacterium sulci, and several others) metabolize proteins and produce volatile sulfur compounds: hydrogen sulfide (the rotten-egg smell), methyl mercaptan, and dimethyl sulfide.

K12 inhibits several of these anaerobes through its bacteriocins, and the lozenge format puts a substantial bacterial load directly on the dorsum of the tongue. The Tagg group's landmark 2005 paper in Journal of Applied Microbiology showed that subjects with confirmed halitosis (measured both organoleptically by trained judges and by gas chromatography for sulfur compounds) saw reductions in volatile sulfur compounds of around 85 percent at one week of K12 use, with the effect maintained over the trial period. Subsequent trials have generally confirmed the direction of the effect, although effect sizes vary.

A 2018 systematic review of probiotics for halitosis (Yoo et al., Journal of Periodontal Research) pooled the available randomized trials and found that probiotic use (K12 in the majority of trials) produced a statistically significant reduction in both organoleptic scores and volatile sulfur compound measurements compared to placebo, with a particularly strong effect when the probiotic was combined with tongue cleaning. The review noted that the trials are heterogeneous and small, but the consistency of the direction of effect across studies is reassuring. The reviewers concluded that oral probiotics are a "reasonable adjunct" to mechanical hygiene for chronic halitosis, especially the morbid-breath kind that persists despite excellent brushing and flossing.

The honest framing here is that for someone with genuine, persistent halitosis (the kind that close partners and friends have actually noted, not the perceived bad-breath anxiety that turns out to be unfounded), K12 is one of the better-evidenced interventions and worth a 30-day trial. For someone whose bad breath is intermittent or situational (morning breath, post-coffee breath, post-onion breath), the gains from K12 are mostly drowned out by simpler interventions like scraping the tongue and rinsing with water. For a deeper map of where bad breath actually comes from, the post on bad breath that won't go away walks through the seven main causes and how to tell them apart.

The M18 story and cavity-causing bacteria

Streptococcus salivarius M18 came out of the same New Zealand screening program in the 2000s. The selection criterion this time was different: instead of inhibiting S. pyogenes, the M18 strain was identified because it strongly inhibited Streptococcus mutans, the bacterium most responsible for the early-stage demineralization that leads to dental caries. M18 produces a different bacteriocin profile (salivaricin M and salivaricin 9) and additionally expresses an enzyme (dextranase) that degrades the extracellular polysaccharide matrix S. mutans uses to build cavity-promoting biofilm.

The mechanism is biologically appealing. S. mutans is a successful cariogen in part because it sticks to enamel using glucan polymers it synthesizes from dietary sucrose. The glucan layer is what makes plaque difficult to remove and what creates the local pockets where acid concentration builds up against the tooth surface. An organism that both kills S. mutans and breaks down its scaffold should, in theory, push the cariogenic balance in a favorable direction.

The clinical evidence is reasonable but not yet conclusive. The most cited trials are from Italian groups (Di Pierro and colleagues, 2014 to 2020) and from the Tagg lab itself. A 2014 trial in 76 children with high S. mutans counts found that 90 days of M18 lozenge use significantly reduced salivary S. mutans counts and plaque acidogenicity compared to placebo. A 2015 trial in adults showed similar reductions. A 2020 review (Probiotics and Antimicrobial Proteins) pooled the available data and concluded that M18 produces a reliable, moderate reduction in cariogenic biomarkers, but cautioned that no trial has yet been large or long enough to show a reduction in actual new cavity rates, which is the clinical outcome that matters most.

This gap (biomarker improvement without proven outcome improvement) is the central limitation of M18's evidence base. The biology is sound, the short-term markers move in the right direction, but the long, expensive cavity-incidence trials have not been done. For high-cavity-risk individuals (orthodontic patients, frequent snackers, dry-mouth users) who already do everything else right, adding M18 is reasonable and low-risk. For people with normal cavity risk, the cost-benefit is more open. Mechanical hygiene, sugar control, and topical remineralizers (fluoride, nano-hydroxyapatite, CPP-ACP) all have stronger outcome evidence than M18 does.

Lactobacillus reuteri and gum inflammation

The third major oral probiotic strain is not a streptococcus at all. Lactobacillus reuteri is a gut-resident lactic acid bacterium that has been studied for general digestive health for decades, but two of its strains (DSM 17938 and ATCC PTA 5289) have a separate body of clinical evidence in oral applications, specifically for gingivitis and mild periodontitis.

The mechanism is partly antimicrobial (L. reuteri produces reuterin, a broad-spectrum compound active against several periodontal pathogens) and partly anti-inflammatory (L. reuteri reduces local cytokine production in gum tissue in cell-culture and animal models). The clinical signal in humans is most consistent in short-term gingivitis trials. A frequently cited 2008 trial (Krasse et al., Swedish Dental Journal) found that 30 days of L. reuteri lozenges produced significantly greater reductions in bleeding on probing and gingival index compared to placebo in subjects with moderate gingivitis. Later trials, including a 2020 systematic review (Cochrane-style meta-analysis published outside Cochrane proper), generally support the direction of effect, although effect sizes are small to moderate.

For established periodontitis (with bone loss and deep pockets), the evidence is much weaker. L. reuteri has been studied as an adjunct to scaling and root planing in moderate periodontitis, with mixed results. The most realistic framing is that L. reuteri is a reasonable addition for someone with persistent gingivitis who already brushes, flosses, and has been to a hygienist. It is not a substitute for periodontal treatment and should not be used as one.

Brands marketing L. reuteri for oral health usually combine it with a streptococcus salivarius strain, on the logic that they target different niches (gum line and tongue, respectively). The combination is reasonable, although the strict per-strain evidence base is for each separately.

How to dose oral probiotics: the boring details that matter

Dosing is where most users get it wrong, and where the gap between trial protocols and consumer use becomes large. The trials that show benefit follow a specific pattern that is easy to replicate at home, and almost as easy to break by accident.

The standard daily dose across well-designed K12 and M18 trials is one to two lozenges per day, each containing at least one billion colony-forming units (CFU) of the named strain at time of consumption. The "at time of consumption" qualifier matters. Probiotic potency degrades during storage; a product that contained 5 billion CFU at manufacture may contain only 1 billion at expiry. Brands that disclose only the manufacture-time count are giving you the optimistic version of the number.

Reading an oral probiotic label without getting fooled

The oral probiotic category is heavily padded with marketing-led products. A quick screen of any label cuts the field in half within seconds.

Beyond the strain name and CFU disclosure, there are a few smaller details worth checking. Sugar-free is non-negotiable: a sweetened probiotic lozenge feeds the same cariogens it is supposed to inhibit. Most reputable products sweeten with xylitol or stevia. The lozenge format should be slow-dissolving, not chewable or hard-cracking; the latter cuts contact time. Storage instructions matter: live bacteria are sensitive to heat, so a brand that ships products in summer without temperature control is gambling with viability.

For comparison shopping, the reputable brands using BLIS K12 or M18 material at the time of writing include Now Foods OralBiotic, Hyperbiotics PRO-Dental, Designs for Health VegeBiotic Throat, ProDentim (which uses a multi-strain blend including K12 and M18 but is more aggressively marketed), and several pharmacy own-brands that white-label BLIS material. The BLIS Technologies direct line (BLIS K12 ThroatGuard, BLIS M18 ToothGuard) is the most strain-pure option and is usually the cheapest per gram of active material.

Where oral probiotics fit in a real routine

The honest verdict on this category is that oral probiotics are a real intervention with a narrow band of usefulness. They are not snake oil, and they are not a foundation of oral health. Sitting them inside a complete routine, in order of evidence-weight, looks roughly like this.

First tier (non-negotiable): mechanical hygiene (brushing twice a day with a soft-bristled brush, flossing or interdental brushing daily, tongue scraping in the morning) and a remineralizing topical (fluoride toothpaste, nano-hydroxyapatite toothpaste, CPP-ACP paste, or a combination). These are the inputs that move outcomes the most across the largest number of users. Without them, no probiotic strategy has anything to attach to.

Second tier (high value for specific people): xylitol delivery (gum or mints, 5 to 10 grams across the day) for anyone with elevated cavity risk, dietary sugar control, and saliva-boosting habits (hydration, sugar-free gum, breathing through the nose). The xylitol and saliva levers are well-evidenced and have the additional advantage that they work for everyone, not just specific microbiome profiles.

Third tier (helpful for specific complaints): oral probiotics. K12 for someone with persistent bad breath that has not responded to mechanical hygiene and tongue scraping. M18 for someone with documented high S. mutans counts or a recent run of cavities despite otherwise good habits. L. reuteri for someone with persistent gingivitis who already does the basics. In all three cases, the probiotic is an adjunct to the first two tiers, not a replacement.

For most healthy adults with standard cavity risk and no chronic halitosis, the cost-benefit of adding daily K12 or M18 is genuinely marginal. The €15 to €30 a month spent on probiotic lozenges might be better invested in dental cleanings, a better electric toothbrush, or a higher-quality remineralizing toothpaste. For the specific groups described above (chronic bad breath, high cavity risk, persistent gingivitis), the same spend can be a reasonable line item.

Safety, side effects, and who should not use them

In healthy adults and children, K12 and M18 have an excellent safety record. BLIS Technologies maintains a post-marketing surveillance database, and across more than 20 years of commercial use in over 80 countries, no serious adverse events have been linked to the strains. Mild gastrointestinal effects (some bloating, occasional looser stools in the first week) occur in a small percentage of users and usually resolve within a few days as the system adjusts.

The well-recognized contraindications for any live probiotic apply. People with severe immunocompromise (HIV with low CD4 counts, active chemotherapy, bone marrow transplant recipients within the first year) should avoid live oral probiotics unless cleared by their treating physician. Patients with central venous catheters carry a small but documented risk of probiotic-related bacteremia, and the same caution applies to people with structural heart disease, prosthetic heart valves, or a history of infective endocarditis. Pregnant and breastfeeding users are generally considered safe to use K12 and M18, although the manufacturer's own labels often suggest a conservative approach of consulting a clinician first.

Drug interactions are minimal. The one consistent recommendation is to space oral probiotics two hours away from any antibiotic dose, both because the antibiotic will kill the probiotic and because the probiotic does not interfere with the antibiotic's primary action. After a course of antibiotics, starting a 30-day K12 protocol can be a useful way to reseed the oral cavity, since broad-spectrum antibiotics tend to thin out the resident flora indiscriminately and leave the door open for opportunistic species.

What the next decade of evidence will probably show

The current oral probiotic evidence base is dominated by short trials with biomarker endpoints. The category is overdue for the kind of large, long-duration trials that have been done for fluoride and gut probiotics. There are several lines of research in active development that will probably change how this guide reads in five years.

First, the personalized-microbiome angle. The same K12 dose produces very different responses in different users, depending on the existing flora. Sequencing-based pre-screening could identify the subset of users who actually benefit from a given strain, instead of treating the population as homogeneous. Several labs (most notably the Joel Kerr group in the UK and groups at NYU and Harvard) are working on this, and the first commercial microbiome-matched recommendations are likely to ship by the late 2020s.

Second, the engineered-probiotic angle. Genetically modified strains of S. mutans itself (the "BCS3-L1" strain developed by Jeffrey Hillman at the University of Florida) were proposed in the early 2000s as a replacement-therapy approach: instead of inhibiting cariogens, replace them with engineered strains that cannot produce acid. The approach has stalled commercially, partly because of regulatory friction around live-modified-organism therapeutics, but the underlying idea is being revisited as gene-editing tools improve.

Third, the combination-product angle. Several companies are working on products that combine oral probiotics with remineralizing minerals (nano-hydroxyapatite or CPP-ACP) and prebiotic substrates (xylitol, arginine). The logic is that addressing multiple levers simultaneously should produce additive or synergistic effects, but the trial data on these combination products is still preliminary. Anyone interested in the ecological logic of how mouth bacteria interact with diet and minerals can read the broader piece on the oral microbiome, which is the foundation that probiotic strategies are built on.

For now, the practical takeaway has not changed in five years and is unlikely to change in the next two. K12 works for halitosis. M18 plausibly helps with cavity-bacteria suppression. L. reuteri can nudge gingivitis. The strains are safe, the dose is modest, and the cost is reasonable. They are not the most important thing you do for your teeth. They are a defensible add-on for specific complaints. Anyone selling them as more than that is overreaching.

Frequently asked questions