Bacterial Vaginosis (BV)

Bacterial vaginosis is among the most misdiagnosed conditions in primary care. Patients arrive describing discharge and a fishy odor, and many have already spent weeks buying over-the-counter yeast treatments that did nothing. That is because BV and yeast are different conditions caused by different organisms, and an antifungal cannot touch a bacterial problem. Knowing the difference, and knowing how to confirm it, saves time, money, and prolonged symptoms.

This guide covers what BV actually is, why it develops, how it is diagnosed, the full treatment menu, and, most importantly, what to do when it keeps coming back. Every clinical claim is cited to a primary source, and all medications and doses reflect the CDC 2021 Sexually Transmitted Infections Treatment Guidelines and the peer-reviewed literature that has appeared since.

What Is Bacterial Vaginosis?

Bacterial vaginosis is a state of vaginal dysbiosis, meaning the normal balance of bacteria in the vagina has shifted in a harmful direction. In a healthy vagina, Lactobacillus species dominate. They ferment glycogen into lactic acid and keep the vaginal pH below 4.5, an acidic environment that discourages most harmful organisms.^[14] In BV, the Lactobacillus population drops sharply, and a mixed group of mostly anaerobic bacteria expands to fill the gap.^[2] This flora is not static; it shifts within the same person over short periods, which is one reason BV can appear and recede over time.^[9]

The bacteria most consistently found in BV include Gardnerella vaginalis, Atopobium vaginae, Prevotella species, and other anaerobes.^[8][30][2] These organisms raise the pH above 4.5, produce the amines behind the characteristic fishy odor, and can form a biofilm that clings to the vaginal wall. BV is therefore not a classic infection with a single pathogen and an immune response. It is more accurately described as an ecological collapse, a loss of the microbial community that normally protects the vagina.^[7]

Because BV is a shift in an ecosystem rather than a fixed infection, the same woman can move in and out of a BV state over weeks without any treatment. The lactobacilli that decline in BV are also the source of the hydrogen peroxide and lactic acid that hold the vaginal pH down and keep pathogens in check. When they fall, the conditions that favor not only Gardnerella but also viral and other bacterial growth improve, which is the mechanistic link between BV and the downstream risks described later in this guide.^[14][7]

What Causes BV and Who Gets It

The precise trigger that sets off the Lactobacillus collapse is not fully understood, but the risk factors are well mapped. A systematic review and meta-analysis found that sexual activity is the dominant and most consistent association: a new male partner, a higher number of partners, and intercourse without condoms all raise risk.^[11] Among women who have sex with women, BV is strongly associated with female partners and shared behaviors, supporting a role for bacterial exchange between partners.^[12]

Other established risk factors from a large U.S. cohort study include the absence of hydrogen peroxide producing lactobacilli, which are protective, and non-Lactobacillus factors such as smoking and a history of herpes simplex virus type 2.^[13] Douching also repeatedly shows up as a risk: a 2007 national survey found BV prevalence climbed with higher douching frequency, and douching was inversely associated with oral contraceptive use, which appears modestly protective.^[4]

This fits a coherent picture. Anything that raises vaginal pH or physically disrupts the protective flora, from semen to menstrual blood to douching, creates the window in which the BV associated anaerobes can establish themselves and, in many women, build a biofilm that makes them hard to fully eradicate.^[28]

The microbiome's own resilience is the other half of the story. Some women carry flora that recovers quickly after a disturbance, while others shift stably into a BV state and stay there, a difference that is only partly explained by behavior.^[9] Population surveys also show that BV prevalence varies by race and ethnicity, with higher rates among Black and Mexican American women even after accounting for sexual activity, a pattern tied to baseline differences in vaginal flora that remain incompletely understood.^[4] These are group patterns, not personal predictions, and they do not change the recommendation to diagnose and treat any symptomatic BV.

Symptoms of BV

The single most important fact about BV symptoms is that most women with BV do not have any. In the National Health and Nutrition Examination Survey, only about 15.7 percent of women with laboratory confirmed BV reported any vaginal symptoms, meaning roughly 84 percent were asymptomatic.^[4] When symptoms do appear, they tend to have a recognizable pattern.

The classic presentation is a thin, homogeneous, gray or off-white discharge that evenly coats the vaginal walls, together with a fishy odor that becomes more noticeable after intercourse or during menstruation.^[5] Itching and irritation are usually mild or absent, which is a useful contrast with yeast infections, where intense itching is the dominant complaint. BV does not cause fever, and it does not cause lower abdominal or pelvic pain by itself, so those findings should prompt a search for something else.

Some women also report a mild burning with urination or irritation at the vaginal opening, symptoms that overlap with a urinary tract infection and with yeast, another reason self-diagnosis misfires. The absence of intense itching is the most consistent negative finding that points away from yeast. Because the discharge is thin and often mistaken for a normal day, many women live with recurrent BV for years before anyone runs the correct test.

For a deeper look at the discharge and odor, see our dedicated guides on BV discharge and BV smell.

How BV Is Diagnosed

You cannot confirm BV from symptoms alone, and clinical impression alone misses a meaningful share of cases. The two standard approaches are the Amsel clinical criteria and the Nugent laboratory score, and newer molecular tests have joined them.

Amsel criteria. Diagnosis requires at least 3 of these 4 findings: a thin, homogeneous vaginal discharge; a vaginal pH above 4.5; the presence of clue cells on saline microscopy, which are vaginal epithelial cells coated with adherent bacteria; and a positive whiff test, meaning a fishy odor released when potassium hydroxide is added to discharge.^[5] Of these, clue cells are the single most specific finding.

Nugent score. This is the laboratory reference standard. A Gram stain of vaginal fluid is scored from 0 to 10 based on the relative abundance of Lactobacillus morphotypes versus Gardnerella and Bacteroides morphotypes and curved rods. A score of 7 to 10 indicates BV, 4 to 6 is intermediate, and 0 to 3 is normal.^[6]

Molecular testing. Nucleic acid amplification tests can now detect the DNA of BV associated organisms directly from a vaginal swab and are increasingly used in clinical and telehealth settings because they do not require microscopy and are highly sensitive.^[1] Molecular diagnosis is a reasonable choice when a clinician cannot perform a wet mount on site. Clue cells themselves have shed new light on the condition, with recent work distinguishing true clue cells, epithelial cells densely coated in Gardnerella, from look-alike pseudo clue cells, a nuance that affects how reliably the finding is read.^[31]

Each method has limits worth knowing. Amsel criteria require an experienced examiner and a microscope, and they can miss cases where only some findings are present. The Nugent score is objective but needs a laboratory and a trained reader, and its intermediate category (a score of 4 to 6) is genuinely ambiguous. Molecular tests are sensitive and convenient but can flag colonization as well as disease, and they cost more. None of this changes the core point: a swab plus a pH check, whether done in an office or through a telehealth-directed self-test, beats guessing from symptoms.

What Has Changed in BV Care: 2021 to 2026

BV management has moved meaningfully since the 2021 CDC guidelines were published, in three directions. First, single-dose treatment became a real option. Secnidazole 2 g, approved by the FDA as a one-time oral granule dose, removes the adherence problem that quietly undermines weeklong courses, at the cost of a somewhat lower cure rate.^[24] Second, the problem of recurrence finally has level one evidence for a new lever: the 2025 partner-treatment trial showed that treating a monogamous male partner cuts 12 week recurrence from 63 percent to 35 percent, the kind of effect size that usually reshapes guidelines.^[23]

Third, non-antibiotic and biofilm-directed strategies moved from theory toward practice. Dequalinium chloride, an antiseptic rather than an antibiotic, met non-inferiority against oral metronidazole in a 2024 trial, an option that sidesteps both antibiotic resistance and common intolerances.^[25] Boric acid, long used off-label, is now better characterized as a biofilm-disrupting support for recurrent disease, with a 2021 safety review consolidating what is safe and what is not, notably the recommendation to avoid it in pregnancy.^[27] And the long-debated warning against drinking alcohol while taking oral metronidazole was revisited in 2026 and found to rest on thinner evidence than generations of caution suggested.^[26]

None of this supplants the fundamentals: the first-line metronidazole and clindamycin regimens remain the backbone, and accurate diagnosis still comes before any prescription. But a patient with recurrent BV in 2026 has real, evidence-based options that simply did not exist in meaningful form a few years earlier.

First-Line Treatment for BV

The 2021 CDC guidelines recommend treating all women with symptomatic BV, both to relieve symptoms and to reduce the downstream risks of STI acquisition and, in pregnancy, obstetric complications.^[1] The three recommended first-line regimens all achieve clinical cure in roughly 80 to 95 percent of cases.^[1][3]

An important practical point about alcohol and oral metronidazole. For years clinicians told patients to avoid alcohol during and for 24 to 48 hours after treatment because of a feared disulfiram like reaction. A 2026 review in Sexually Transmitted Diseases re-examined the evidence and found the reaction is not reliably supported by data.^[26] The CDC guidance still lists alcohol avoidance as a caution, so individual clinicians may reasonably advise either way, but the widely repeated warning is now contested rather than settled science.^[1]

Finish the entire course even if symptoms improve after a day or two. Stopping early suppresses but does not clear the bacterial population, and that is one of the reasons BV returns. Avoid douching at any point, and consider avoiding intercourse or using condoms during treatment so that semen and a partner's microbiome do not work against the antibiotic.^[1]

Because metronidazole is the most used first-line option, its specifics deserve their own page. For detailed dosing, the alcohol caution in full, and side effect management, see our metronidazole guide.

Alternative and Adjunctive Regimens

When a first-line option is contraindicated or poorly tolerated, the CDC recognizes several alternatives: clindamycin 300 mg orally twice daily for 7 days, clindamycin 100 mg ovules inserted at bedtime for 3 days, secnidazole 2 g as a single oral dose, or tinidazole, at 2 g once daily for 2 days or 1 g once daily for 5 days.^[1]

Secnidazole deserves a closer look because of its convenience. It is an oral 5-nitroimidazole taken once as granules sprinkled on soft food, and an integrated analysis of two registration trials found it superior to placebo on clinical cure, with the single-dose advantage of eliminating adherence problems.^[24] Its tradeoff is a somewhat lower headline cure rate than a full course of metronidazole, a point separated out in the chart above.

An emerging non-antibiotic option is dequalinium chloride, a broad-spectrum antiseptic. A 2024 randomized trial compared dequalinium chloride vaginal tablets to oral metronidazole and met its non-inferiority margin, with a clinical cure rate of 92.8 percent in the dequalinium arm at the first visit.^[25] Dequalinium chloride is not yet a U.S. mainstay but illustrates the direction of allergy friendly and resistance sparing options.

Choosing a Treatment: A Decision Framework

All of the regimens above work, but the best choice depends on the scenario. The table maps common situations to a reasonable default, with the understanding that a clinician will individualize based on allergies, preferences, and history.

The one choice that is rarely correct is doing nothing while symptoms persist, or self-treating with an antifungal on a guess. Both extend the episode and, in recurrent cases, let the biofilm reseed. When in doubt, confirm the diagnosis first, then treat.

Why BV Keeps Coming Back

Initial treatment works well, but the biology of BV makes recurrence the rule rather than the exception. A cohort study that followed women for a full year after oral metronidazole found that 58 percent had BV return within 12 months, and 69 percent had return of abnormal vaginal flora.^[10]

Two mechanisms drive this. First, antibiotics suppress the BV associated bacteria but do not restore Lactobacillus. When the antibiotic clears, the ecological conditions that allowed BV in the first place are often still present, so the anaerobes repopulate.^[7][18] Second, Gardnerella and related organisms form a biofilm on the vaginal epithelium that antibiotics penetrate poorly. That biofilm can act as a reservoir, reseeding the vagina even after an apparently successful treatment.^[28] Swidsinski and colleagues showed the same Gardnerella biofilm is found on male partners and appears to be exchanged sexually, which helps explain the partner treatment finding discussed below.^[29]

For a full breakdown of recurrence and how to stop the cycle, see how to get rid of BV.

Recurrent BV: Suppressive Therapy

For women who get BV three or more times a year, a single repeated antibiotic course is usually not a durable answer. The strongest evidence supports suppressive therapy with metronidazole gel. In a randomized trial, women who began with a 10 day metronidazole gel course and then, if cured, used twice weekly metronidazole gel for 16 weeks had a recurrence of only 25.5 percent, versus 59.1 percent on placebo.^[19]

Why suppression works when single courses fail reflects the biofilm problem described above. A short antibiotic course knocks down free-floating bacteria, but the biofilm-bound reservoir survives, so the stage is set for rapid regrowth the moment treatment stops.^[28] Suppression instead applies a low, sustained dose that holds the anaerobes in check while the normal flora is given time to reestablish. The caveat is that protection fades once suppression stops: in the same trial, recurrence by 28 weeks, which included 12 weeks off therapy, rose to 51 percent on metronidazole versus 75 percent on placebo, so suppression is a maintenance strategy, not a permanent cure.^[19]

A related strategy adds boric acid to the induction phase. In a retrospective series, women with recurrent BV treated with 7 days of oral nitroimidazole followed by 21 days of intravaginal boric acid 600 mg had cure rates of 88 to 92 percent at 7 to 12 weeks, with cumulative cure of 87 percent, 78 percent, and 65 percent at 12, 16, and 28 weeks respectively.^[20] Boric acid is thought to act by disrupting biofilm, targeting the mechanism that simple antibiotics miss.^[20][28] Boric acid is covered in depth in our boric acid for BV guide.

A combined long term regimen described by Surapaneni and colleagues weds an oral nitroimidazole, boric acid, and metronidazole gel into a staged plan and, in their experience, achieved extended remission in most treated women.^[21] The live biotherapeutic approach is also active. A 2020 randomized trial tested Lactobacillus crispatus CTV-05 (Lactin-V), applied vaginally after standard metronidazole treatment, and found recurrence at 12 weeks in 30 percent of the Lactin-V group versus 45 percent on placebo.^[22] Lactin-V is not yet commercially available as a prescription in the U.S., but it signals where treatment may be headed.

The 2025 Male Partner Treatment Trial

The most practice-changing BV finding in years came from a 2025 New England Journal of Medicine trial. Couples in which the woman had BV were randomized so that either the woman alone received standard treatment, or the male partner was also treated with oral metronidazole plus clindamycin cream applied to the penile skin for 7 days. At 12 weeks, BV had recurred in 35 percent of women whose partners were treated versus 63 percent of those whose partners were not.^[23] The trial's safety board stopped it early because treating the woman alone was clearly inferior.

This does not yet change the 2021 CDC recommendation, which does not advise routine male partner treatment, but it is strong evidence that the next guideline update may, and it gives clinicians and couples in monogamous relationships a data backed option to discuss for highly recurrent BV.^[1][23] The biology makes sense: the Gardnerella biofilm found on men is the same biofilm implicated in recurrence among women.^[29]

BV vs Yeast Infection: How to Tell the Difference

This is the question clinicians hear most, because the two conditions are superficially similar but treated with opposite medication classes. BV is a bacterial overgrowth treated with antibiotics; yeast (vulvovaginal candidiasis) is a fungal overgrowth, usually Candida albicans, treated with antifungals.^[32] The single most reliable bedside discriminator is vaginal pH, which is elevated in BV and normal in yeast, but discharge, odor, and itching point in different directions too, as shown in the table in the symptoms section above.

The practical takeaway: a strong fishy odor with thin discharge and little itching points to BV, while intense itching with thick white discharge and no odor points to yeast. The two can occur together, especially after antibiotics for BV trigger a secondary yeast overgrowth. Treating the wrong one does nothing, which is why self-treatment with over-the-counter antifungals so often fails. Get a diagnosis before treating. The full differential is in our BV vs yeast infection guide, and the yeast side belongs to the yeast infection pillar.

BV in Pregnancy

BV during pregnancy matters because it is associated with preterm birth. A 2023 meta-analysis of 20 studies and nearly 300,000 observations found BV raised the odds of preterm birth roughly twofold (odds ratio 1.79), with a relative risk of 1.44.^[16] The relationship appears strongest when BV is present early or persists.^[16]

The treatment approach in pregnancy follows the same antibiotic classes but with timing considerations. The Society of Obstetricians and Gynaecologists of Canada screening and management guideline describes treating symptomatic BV in pregnancy and notes that oral metronidazole is generally avoided in the first trimester where possible.^[17] Pregnant patients should have any suspected BV confirmed and managed by their obstetric care team rather than self-managing, because boric acid in particular is not recommended in pregnancy.^[27]

BV and Risk of HIV and Other STIs

The consequences of untreated BV extend well beyond symptoms. A meta-analysis found that BV is associated with a significantly higher risk of HIV acquisition, with a pooled odds ratio of about 1.6.^[15] The likely mechanisms include disruption of the acidic protective barrier, the loss of hydrogen peroxide producing lactobacilli, and vaginal inflammation.^[13][15] The same disruption is linked to higher susceptibility to other STIs such as chlamydia, gonorrhea, and trichomoniasis, and to pelvic inflammatory disease.^[1] This is why even a symptom free, incidentally discovered BV diagnosis in a high-risk setting can be worth treating, a judgment a clinician should make with you.^[1]

BV and Daily Life

BV is not dangerous in the way a high fever is, but its day to day toll is real and underappreciated. In qualitative research, women with BV routinely describe feeling unclean, avoiding sex, and worrying that a partner will interpret the odor as poor hygiene, and many go through repeated cycles of home remedies and over-the-counter products before anyone makes the correct diagnosis.^[33] The condition lands squarely at the intersection of physical symptom and self-image, which is why it feels heavier than a simple infection.

The misdiagnosis loop is the most damaging pattern in practice. A woman notices discharge, assumes yeast, buys an antifungal that does not touch bacteria, and returns weeks later more frustrated and convinced something is seriously wrong.^[32] Then the recurrence cycle compounds it: with 58 percent of women having BV return within a year, each new episode resets the clock on intimacy and confidence.^[10] The honest message is that none of this is the patient's fault, BV is common, and a correct diagnosis paired with a realistic recurrence plan takes most of the weight off.

For many women, simply having the condition named and explained is itself a relief, because it reframes the odor and discharge from a hygiene failure into a recognizable, treatable medical state.^[33] If the emotional impact of recurrent BV is affecting you, that is a legitimate reason to seek care, and it is one your clinician should take seriously rather than wave off.

Prevention

There is no sure way to prevent BV, but the risk factors point to a few evidence supported steps. Avoiding douching is the clearest, since douching strips the protective lactobacilli and disrupts pH, and no medical organization recommends it for any reason.^[4] Consistent condom use buffers semen's alkaline effect and reduces BV risk, and it removed the main sexual risk factor in the trials.^[11][13] Hormonal contraception, particularly oral contraceptives, is associated with lower BV in multiple studies, likely by stabilizing the vaginal environment.^[4][13]

One caveat about supplements: commercial oral and vaginal probiotics sold for women's health are not interchangeable with the specific Lactobacillus crispatus strain tested in trials, and the CDC does not recommend them for prevention.^[1] Hormonal status also matters, since BV is more common during estrogen dips such as the postpartum period and, in some women, in the days before a period, a reflection of how tightly the vaginal microbiome tracks hormones.^[4]

For women in a monogamous relationship with highly recurrent BV, partner treatment is now a legitimate, trial-supported strategy to discuss with a clinician, even if it is not yet in the guideline.^[23] More broadly, treating the condition decisively rather than ignoring a mild case matters, because leaving BV untreated keeps the door open to the STI and pregnancy risks described below, and repeated untreated episodes tend to seed the biofilm reservoir that makes future treatment harder.^[15][28]

What you cannot reliably prevent with supplements is covered next, but if BV does recur, the recurrences are manageable with the suppression and partner strategies described above, not with endless single courses.

What Does Not Work for BV

Several common approaches lack evidence or are actively harmful. Over-the-counter antifungal creams and suppositories do nothing for BV because they target yeast, not bacteria, and this is the single most common reason symptoms do not resolve.^[32] Douching, whether for hygiene or as a home remedy, worsens the very imbalance it is meant to fix.^[4]

Probiotics by mouth or vagina have shown mixed results, and the CDC does not currently recommend them as either adjunctive or replacement therapy.^[1] The exception is the specific strain Lactobacillus crispatus in the Lactin-V trial, which did reduce recurrence, but that is an investigational live biotherapeutic, not a retail supplement.^[22] Garlic cloves, yogurt inserts, tea tree oil, and hydrogen peroxide douches are not supported by evidence and can irritate the mucosa. If a home remedy sounds too good to be true, it usually is.

Red Flags: When to Contact a Clinician

BV itself is not an emergency, but several findings should prompt prompt evaluation rather than watchful waiting. Contact a clinician if you have fever, lower abdominal or pelvic pain, pain during sex, or abnormal uterine bleeding, which can signal pelvic inflammatory disease or another infection.^[1] Likewise, any symptom during pregnancy, any discharge with blood, or symptoms that do not improve after a completed antibiotic course warrant a revisit. Finally, if a treatment course fails more than once in a short span, the working diagnosis may not be BV at all, and a trichomoniasis or other vaginitis workup is indicated.^[1]

Frequently Asked Questions