Bacterial vaginosis is diagnosed through a combination of symptoms, a vaginal pH test, microscopic examination, and sometimes molecular testing. There’s no single test that confirms it on its own. A clinician typically needs at least three out of four specific clinical signs to make the diagnosis, or a lab can score a vaginal sample under a microscope. Self-diagnosis is unreliable: studies show women correctly identify BV only about 56% of the time based on symptoms alone.
The Four Clinical Signs Clinicians Look For
The most widely used method for diagnosing BV in a clinical setting is called the Amsel criteria. Your provider checks for four specific signs during a pelvic exam, and at least three must be present to confirm BV:
- Thin, milky discharge that smoothly coats the vaginal walls, rather than clumpy or thick discharge
- Vaginal pH above 4.5, tested with a small strip of pH paper touched to the vaginal wall
- A fishy smell when a drop of potassium hydroxide solution is added to a sample of the discharge (known as the “whiff test”)
- Clue cells visible under a microscope, which are vaginal skin cells covered in a layer of bacteria that obscures their edges
Of these four, the clue cell finding is generally considered the most reliable single indicator. When at least 20% of the vaginal cells on a slide are clue cells, it strongly supports a BV diagnosis. The pH test, on the other hand, is less specific. Vaginal pH above 4.5 can also occur with other infections, during menstruation, or after unprotected sex, so it’s never used as the sole deciding factor.
Lab-Based Microscopy Scoring
The reference standard for diagnosing BV in a laboratory setting is a scoring system applied to a Gram-stained vaginal sample. A lab technician examines the slide under a microscope and assigns a score from 0 to 10 based on the types and quantities of bacteria present. A score of 0 to 3 means the sample is dominated by healthy lactobacilli, which is normal. A score of 4 to 6 falls into an intermediate zone where BV-associated bacteria are starting to emerge but haven’t fully taken over. A score of 7 to 10 confirms BV, showing that lactobacilli have been largely replaced by the anaerobic bacteria characteristic of the condition.
This scoring method is more objective than the bedside clinical criteria because it relies on standardized bacterial counts rather than a clinician’s judgment about discharge consistency or smell. However, it requires a lab and trained technician, so results aren’t instant. Many clinics use the bedside Amsel criteria for an immediate answer and reserve the lab scoring for ambiguous cases or research settings.
Molecular and DNA-Based Tests
Newer molecular tests can detect the DNA of specific bacteria associated with BV from a vaginal swab. These tests measure both the BV-causing bacteria and the protective lactobacilli, then use an algorithm to determine whether the bacterial balance has shifted enough to qualify as BV. Results typically come back within 24 hours, and some of these tests can be performed on self-collected swabs rather than requiring a speculum exam.
These molecular panels often test for BV, yeast infections, and trichomoniasis simultaneously from a single sample, which is useful because the symptoms of all three overlap considerably. They have high sensitivity and specificity for BV in women who are already experiencing symptoms like discharge, odor, or itching. However, their accuracy hasn’t been well established for women without symptoms, so they’re currently recommended only for symptomatic testing.
What Doesn’t Work for Diagnosis
A standard bacterial culture is not useful for diagnosing BV. The primary bacterium involved, Gardnerella vaginalis, can be found in women without BV, so growing it in a culture doesn’t prove anything. Pap smears are similarly unreliable for this purpose, with low sensitivity and specificity for detecting BV. If a Pap smear incidentally mentions BV-associated changes, it may warrant follow-up, but a Pap test is not a diagnostic tool for this condition.
Why Self-Diagnosis Is Unreliable
Many women assume they can tell the difference between BV and a yeast infection based on symptoms, but the data suggests otherwise. In a study of over 500 women with vaginal symptoms, self-diagnosis of BV was accurate only 56% of the time. When women believed they had a “bacterial infection,” they were correct about 64% of the time, but when they believed they didn’t have one, they were wrong nearly 70% of the time. That means a large number of women who think they have a yeast infection actually have BV, and vice versa.
The practical consequence is significant. About 20% of women who self-diagnosed would have treated the wrong infection entirely, and roughly 24% would have missed a BV diagnosis that needed treatment. Since BV is treated with antibiotics and yeast infections with antifungals, treating the wrong condition won’t help and may make things worse.
Over-the-Counter Screening Kits
Home vaginal pH test kits have been available since 2001. You insert a small swab, then compare the color change on the test strip to a chart. If your pH is above 4.5, the test directs you to see a provider. The problem is that this test is only about 73% sensitive and 67% specific for BV. An elevated pH can also mean trichomoniasis or simply reflect recent sexual activity, menstruation, or douching.
A second type of home test detects an enzyme called sialidase, which is produced by BV-associated bacteria. This test is more specific (about 90%), meaning if it’s positive, you very likely have BV. But its sensitivity is only around 40%, so it misses more than half of actual BV cases. A negative result on this test doesn’t rule BV out.
Neither home test replaces a clinical evaluation. They can be useful as a screening step to help you decide whether to schedule an appointment, but they can’t confirm or rule out a diagnosis on their own.
How BV Is Distinguished From Other Infections
BV, yeast infections, and trichomoniasis can all cause abnormal discharge and discomfort, which is why a clinical exam matters. Several features help a provider tell them apart during the same office visit.
BV typically produces a thin, grayish-white discharge with a fishy odor and minimal inflammation or itching. The vaginal walls usually look normal, not red or swollen. A yeast infection, by contrast, causes thick, clumpy white discharge (often described as cottage cheese-like) with significant itching, redness, and swelling, but the pH is usually normal and there’s no fishy smell. Trichomoniasis can produce a frothy, yellow-green discharge with a strong odor and noticeable irritation, and it’s caused by a parasite that’s visible as a moving organism under the microscope.
When a clinician examines a wet mount slide, the presence of clue cells points to BV, motile parasites point to trichomoniasis, and branching fungal structures point to yeast. If white blood cells are abundant but none of these organisms appear, the provider may investigate cervical infections instead. This layered approach, using pH, microscopy, and the whiff test together, is why an in-office evaluation remains the most reliable path to an accurate BV diagnosis.