Every culture is named, and its dose published. We reviewed nine of the best-selling women's live-culture products on sale in the UK and the US. One published a dose for each individual culture, and it is not sold here. Three published neither a dose nor a strain designation for any of them.

Every culture sits at or above the level used in published research. Trials that found an effect generally used between one and ten billion of a strain per day. The smallest inclusion in this formula is two billion. A low total spread thinly across several cultures can leave the individual doses below anything that has been tested — and without a breakdown, there is no way for a buyer to tell. A trademarked blend name makes that worse rather than better: where a total is published under one, the split between the cultures is not disclosed at all.

The evidence grade is published, including where it is poor. We have not found another product in this category that grades its own ingredients at all.

Most products in this category contain no Bifidobacterium. This one has two. And L. crispatus — the species associated with the most stable vaginal community — takes the largest single share of the formula, at 14 billion.

Trials that found an effect generally used between one and ten billion CFU of a strain per day. Below roughly one billion, a strain is unlikely to be doing anything. Above ten billion, the evidence for additional benefit runs out — and the largest trial in this field found one billion and ten billion performing about the same.

There is also no convincing evidence that a blend of many strains outperforms one or two. The most citable trials in this field are single- and dual-strain products.

Which makes the useful question not how many cultures in total, but is each individual culture present at a dose the research actually used. That is the number we think is worth printing, so here it is.

Grades here are decided largely by who owns a strain, not by what is in the bottle.

An A requires multiple human trials on one exact strain. Those trials cost six figures, and a company funds one only where it can own the result — so the strains that carry A grades are trademarked, and supplied only by the companies that paid for the research, on their terms. A small number of products use them, and those products have earned what they say. Most manufacturers cannot get those strains at all.

Everything else in this category sits at C or below: the same widely available species, from the same suppliers, without individual trials behind them.

So a C is not a poor result against this field. It is the ordinary one, and it is where most of what is on sale would land if anyone else published a grade. What separates one C from another is everything the grade does not measure.

What the grade does not measure

It tells you how much published research exists. It tells you nothing about:

• The dose — whether each culture is present in an amount the research actually used.
• Whether you are told the dose — a single total can hide any distribution beneath it.
• What may lawfully be claimed — and for live cultures in Great Britain, the answer is nothing at all.

Those are separate questions from the grade. The dose and the published counts are set out above; what may be claimed is set out below.

The GR-1 and RC-14 work that forms the backbone of the positive case is bound up with the commercial lineage of those strains, licensed through large ingredient houses.

The strongest "oral probiotics don't reach the vagina" messaging comes from companies selling a vaginally administered product, and rests on a comparator trial that is company-reported and presented at a conference rather than fully peer reviewed.

Neither side is disinterested, and nor are we. We sell the product this page is about. That is the reason for publishing the counter-evidence alongside everything else rather than waiting to be asked for it.

When reviewers assess a field they rate two different things: which way the findings point, and how much confidence to place in them. Those can come apart. A set of small, varied studies can point consistently in one direction while still being too weak to rely on. That is where this field sits.

Where the results point

More trials in this field have reported a difference than have reported none, and where results have been pooled the combined estimate points the same way — most clearly for bacterial vaginosis when live cultures are taken alongside antibiotic treatment.

Not every trial agrees. One of the larger ones found no advantage at all, and it is set out in full further up rather than left out. But the disagreement between trials is mostly about the size of the effect, not about which direction it runs in.

That distinction matters for what follows. The confidence rating is low because the trials are small and varied, not because they contradict each other about whether there is an effect.

What the pooled analyses show

A 2026 systematic review in the American Journal of Obstetrics and Gynecology pooled trials that added live cultures to antifungal treatment and measured recurrence at six months. The pooled estimate favoured the culture groups. The same review found no clear benefit over the longer term.

Two things about it matter more than the headline result does.

Its confidence interval is very wide, meaning the size of any real effect is highly uncertain, and the reviewers rated the finding low certainty.

And the review does not separate cultures that are swallowed from cultures applied vaginally. The authors planned that comparison and could not run it, because there were not enough trials. So the figure describes live cultures in general, not a swallowed capsule specifically.

What the formal ratings say

• The Cochrane review of live cultures for bacterial vaginosis found insufficient evidence to recommend them.
• The Cochrane review covering vulvovaginal candidiasis rated the evidence low to very low quality.
• Later meta-analyses do report a benefit when live cultures are taken alongside antibiotics, but with wide variation between trials.

What those phrases mean

"Insufficient evidence" means there are not enough good trials to judge — not that something has been tested and shown not to work. Those two get read as the same thing and they are not remotely the same thing.

"Low quality" is not a verdict on how carefully any single trial was run. It is a rating applied to a whole body of evidence, and it is marked down when the trials are small, short, inconsistent with one another, or funded by a party with an interest in the outcome. This field scores badly on all four. The studies behind it typically run from a few dozen to a few hundred people, over a few weeks, on different strains at different doses in different groups of women.

Where that leaves the phrase "clinically studied"

It is worth being straight about what this does and does not do for a product like ours.

It does not make an untested culture the equal of a tested one. A small, short trial is still a great deal more than no trial at all, and we are not going to pretend otherwise — it is why we grade our own cultures C rather than something more flattering.

What it does mean is that "clinically studied" is a weaker signal than the phrase makes it sound. In this category it rarely describes a settled question. More often it describes a single modest trial that the reviewers assessing it have rated low quality. Which is the reason this page sets out the studies themselves, their size and their limits, rather than resting on the adjective.

• That this product colonises, restores, repopulates or rebalances the vaginal microbiome.
• That evidence produced on GR-1, RC-14, BB-12, CTV-05, CECT 30648, BB536, MG989 or MG901 applies to the strains in this formula. It does not.
• That any Bifidobacterium in this formula does anything for vaginal health.
• That a higher total culture count is better than a lower one.
• That this product treats, prevents or cures any condition. Live cultures carry no authorised health claim in Great Britain, and we make none.