Bacterial vaginosis is diagnosed through a combination of physical exam findings, vaginal pH testing, and microscopic analysis of vaginal fluid. In most clinical settings, a provider uses a set of four bedside criteria and needs at least three of them to confirm a diagnosis. The process is straightforward and typically happens during a single office visit.
The Four Clinical Criteria
The standard method for diagnosing BV in a clinical setting is called the Amsel criteria, a checklist of four specific findings. At least three of the four must be present for a positive diagnosis:
- Thin, homogeneous discharge: The vaginal discharge has a uniform, smooth consistency and is typically gray-white to yellowish. This differs from the clumpy discharge of a yeast infection or the frothy, greenish discharge more common with trichomoniasis.
- Elevated vaginal pH: Normal vaginal pH sits below 4.5, kept acidic by healthy bacteria. In BV, that pH rises above 4.5 as the balance of bacteria shifts. Your provider measures this by touching a pH strip to the vaginal wall.
- Positive whiff test: A drop of potassium hydroxide solution is added to a sample of vaginal fluid on a glass slide. If BV is present, this releases a distinct fishy odor. Importantly, simply noticing a fishy smell on your own does not count as a positive whiff test. The chemical reaction with potassium hydroxide is what makes it a reliable diagnostic marker.
- Clue cells on microscopy: Your provider examines a sample of vaginal fluid under a microscope. Clue cells are vaginal lining cells so heavily coated with bacteria that their edges appear blurred or stippled. When 20% or more of the visible cells are clue cells, this criterion is met.
No single criterion alone is enough. Vaginal pH, for instance, can also be elevated by semen, menstrual blood, or other infections. The strength of the Amsel approach is that combining multiple markers makes the diagnosis more reliable than any one finding.
What Happens During the Exam
The diagnostic process is quick. During a pelvic exam, your provider collects a small sample of vaginal discharge using a swab. That sample goes onto a glass slide for the whiff test and microscopic examination. A separate pH strip is used to check acidity. Most of the time, your provider can tell you the result before you leave the office.
Under the microscope, your provider is looking at more than just clue cells. The overall landscape of bacteria matters. A healthy sample shows clusters of normal-looking vaginal lining cells and a background dominated by rod-shaped bacteria (the beneficial kind). In BV, those rods are largely replaced by dense collections of smaller, round bacteria clinging to cell surfaces. The presence of curved rods in the sample is another indicator that points specifically toward BV rather than other conditions.
Lab-Based Scoring for Research Settings
A more detailed method called the Nugent score is sometimes used, particularly in research. This involves staining a vaginal fluid sample and scoring it on a scale from 0 to 10 based on the types and quantities of bacteria visible. Scores of 0 to 3 are considered normal, 4 to 6 are intermediate, and 7 to 10 indicate BV. This method is highly standardized and reproducible, which makes it valuable for studies, but it requires lab processing and isn’t as practical for a routine office visit.
Molecular Tests
Newer molecular tests can detect BV by identifying the DNA or RNA of specific bacteria associated with the condition. One FDA-cleared test, the Aptima BV assay, measures levels of both the harmful bacteria that cause BV and the protective bacteria that should be present. In clinical trials, this test was 95% to 97% sensitive, meaning it correctly identified BV in the vast majority of people who had it, with specificity around 86% to 90%.
One advantage of molecular testing is that it works with self-collected swabs just as well as, and in some cases slightly better than, clinician-collected ones. This opens the door for testing through telehealth visits or mail-in kits from your provider. These tests are typically sent to a lab, so results take a day or two rather than being available immediately.
At-Home pH Tests
Over-the-counter vaginal pH test strips are available at most pharmacies. They can tell you whether your vaginal pH is elevated, but they cannot diagnose BV on their own. In a study comparing self-testing to clinical diagnosis, self-collected pH strips were 73% sensitive and 67% specific. That means they miss about one in four cases and flag a fair number of people who don’t actually have BV.
Part of the problem is subjectivity. These tests require you to match a color change on the strip, from yellow to green, which can be hard to interpret. An elevated pH also doesn’t distinguish between BV, trichomoniasis, and other causes of vaginal irritation. A pH strip can be a useful signal that something is off, but it’s not a substitute for a clinical evaluation that includes microscopy or molecular testing.
How BV Is Told Apart From Other Infections
BV, yeast infections, and trichomoniasis can all cause abnormal discharge, so telling them apart matters for getting the right treatment. The characteristics of the discharge itself are only modestly helpful in distinguishing BV from trichomoniasis, since both involve overgrowth of anaerobic bacteria and share overlapping features. Microscopy is the most powerful tool for separating the two.
Trichomoniasis is caused by a parasite that can sometimes be spotted moving on a wet mount slide. It also tends to cause visible inflammation of the vaginal walls and cervix, along with pain during sex, neither of which is typical of BV. BV, by contrast, is not an inflammatory condition. The vaginal walls usually look normal, and the primary symptoms are discharge and odor rather than pain or redness.
Yeast infections are easier to distinguish. They produce a thick, white, cottage cheese-like discharge, vaginal pH stays in the normal range below 4.5, and microscopy reveals yeast cells or branching filaments rather than clue cells. If your pH is normal, BV is effectively ruled out.