Something else? Ask Aunt Vadge – bottom left of your screen!
If you have symptoms of bacterial vaginosis (BV), the most common way it is diagnosed in the room is Amsel’s criteria – a quick, low-cost set of four bedside checks a clinician can do during your appointment, without sending anything to a lab. The four are the look of your discharge, your vaginal pH, a fishy smell when a drop of potassium hydroxide is added, and clue cells seen under a microscope. If three of those four are positive, you have a BV diagnosis.1
It is fast, it is cheap, and decades on it is still one of the most widely used BV tests in the world. But it has real blind spots, and newer molecular testing is changing what a thorough diagnosis can look like. Here is how it actually works, how good it is, and what it can and cannot tell you about your vagina.
How is bacterial vaginosis diagnosed?
Bacterial vaginosis is diagnosed by examining a sample of vaginal discharge. A clinician can check it at the bedside using Amsel’s criteria, a lab can read a Gram-stained slide using the Nugent score, or a molecular test can detect the DNA of BV-associated bacteria. Amsel’s criteria is the fastest and most widely used in everyday practice, which is why it is what most people meet first.1
What is Amsel’s criteria?
Amsel’s criteria is a set of four clinical signs used to diagnose BV at the point of care – meaning in your doctor’s office, during the appointment, rather than days later from a lab report. It was first published by Amsel and colleagues in 1983, and built on earlier methods of reading vaginal samples.2
The appeal is simple. It needs only a vaginal swab, a microscope and slide, some potassium hydroxide (KOH) solution, and pH paper. No expensive machines, no specialist lab, no waiting. That makes it useful almost anywhere, from a busy city clinic to a remote health centre.1
The four Amsel criteria, explained
A diagnosis of BV is made when at least three of the following four signs are present.1 Here is what each one is actually measuring, in plain language.
1. Thin, even discharge
The first sign is a thin, smooth, homogenous discharge – it all looks the same throughout, rather than clumpy or curd-like. The colour is usually grey to off-white, and the amount varies from person to person. This even, milky look is quite different from the thick, cottage-cheese discharge of a yeast infection.1
2. Vaginal pH above 4.5
A protective vagina is acidic, sitting at a vaginal pH of roughly 3.8 to 4.5. In BV, the protective lactic-acid-producing bacteria are crowded out, the pH climbs above 4.5, and the environment becomes friendlier to disruptive bacteria.1
This is the most sensitive of the four criteria, but also the least specific – plenty of other things lift vaginal pH, including your period, recent sex, and lower oestrogen around menopause. A raised pH on its own does not mean BV.3
3. A positive whiff test
The whiff test checks for a telltale fishy smell. A drop of 10 per cent KOH solution is added to a sample of vaginal fluid, and if BV is present it releases a distinctive fishy odour from the amines produced by disruptive bacteria.1
One important detail: the whiff test only counts as positive after the KOH is added by a clinician. Telling your doctor that your discharge smells fishy is a useful clue, but it is not the same as a positive whiff test on its own.1
4. Clue cells under the microscope
Clue cells are vaginal skin cells so heavily coated in bacteria that their edges look fuzzy and speckled rather than crisp. They are seen on a wet mount – a drop of vaginal fluid examined under a microscope. When clue cells make up more than 20 per cent of the cells in the sample, this criterion is met.1
Clue cells are the single strongest sign on the list. One study found that clue cells alone were about 98 per cent specific and 90 per cent sensitive for BV – which is why some clinicians lean heavily on them.4
What to expect at your BV test
The test itself is quick and much like having a cervical screening. You lie back, a clinician gently inserts a speculum and takes a small swab of vaginal fluid. It can feel a little uncomfortable if your tissues are tender, but it is over in moments.5
To keep the result accurate, it helps to avoid sex, douching and any vaginal washes, sprays or creams for about 24 hours beforehand, because these can shift your pH and wash away the very clues the test looks for.5
If your clinician uses Amsel’s criteria, you can often get an answer in the same appointment. If the sample is sent off for a Nugent score or a molecular test, results usually take a few days to come back.5
Amsel’s criteria versus the Nugent score
The other classic BV test is the Nugent score, long considered the laboratory gold standard. It involves staining a vaginal smear and counting three types of bacteria under oil-immersion microscopy, then adding up a score from 0 to 10. A score of 7 or more means BV; 0 to 3 is negative, and 4 to 6 sits in between.6
Nugent is reliable and reproducible, but it is slow and needs a skilled microscopist, so it lives mostly in research and reference labs rather than everyday clinics.6 Amsel’s criteria, by contrast, can be done while you wait.1
Reassuringly, the two largely agree. Studies comparing Amsel’s criteria against Nugent as the reference repeatedly find good concordance, which is why Amsel has comfortably stayed in clinical use for symptomatic BV.37 What matters more than the choice between them is whether the test is done properly and read by someone confident with a microscope.
How accurate is Amsel’s criteria, really?
Amsel’s criteria is strong on specificity – when it says you have BV, it is usually right. Specificity is often reported around 90 per cent or higher, so false positives are relatively uncommon.37
High specificity has a flip side worth knowing about: a positive result tells you the bacterial picture looks like BV, not that you necessarily feel unwell. Plenty of people meet the criteria with no symptoms at all, so it is worth understanding whether BV without symptoms actually needs treating.
Sensitivity is the weaker side, and it is far more variable. Reported sensitivity ranges widely between studies, with some as low as 37 per cent, meaning a genuine case of BV can slip through with fewer than three positive criteria.1 This matters most for people whose symptoms do not read as a textbook case.
There is also a ‘modified’ version that diagnoses BV on just two of the criteria, usually leaning on clue cells plus pH or discharge, and dropping the sometimes-unpleasant whiff test. This can pick up more cases, and some combinations of two positive criteria reach very high specificity – but the modified approach is not a single standardised definition, so it varies from clinic to clinic.1
The criteria are also less reliable around menopause, when lower oestrogen naturally raises vaginal pH and can muddy the reading.1
Where molecular testing fits now
Newer molecular tests – nucleic acid amplification tests, or NAATs – detect the DNA of specific BV-associated bacteria. They are objective, do not need a microscope or a skilled reader, can pick up bacteria at low levels, and work well on a swab you collect yourself.8
For all that, the major guidelines still rate Amsel’s criteria and Nugent as useful for diagnosing symptomatic BV, because they are cheap and fast. Molecular panels are not yet a single new gold standard, and a positive DNA result in someone with no symptoms is not automatically something to treat.8
Molecular and sequencing-based testing comes into its own in the harder cases: recurrent BV, mixed pictures, or symptoms that never quite fit. A bedside test gives you a yes-or-no on BV; a thorough microbiome test shows you which bacteria are actually there and in what proportions. The two answer different questions, which is partly why different vaginal tests can disagree without either being wrong.
What this means for your vagina
Amsel’s criteria is really a snapshot of your vaginal ecosystem at one moment. Each sign – the thin discharge, the raised pH, the fishy whiff, the clue cells – is a different fingerprint of the same underlying shift: protective, acid-making bacteria losing ground to a mix of disruptive bacteria that often build a stubborn biofilm on the vaginal wall.1
That is also why it can be confused with other conditions. Aerobic vaginitis can raise vaginal pH too but has a very different bacterial cast and inflammatory pattern, and the bedside criteria are not built to tell them apart – which is where the differences between AV and BV really matter for treatment.1
Amsel’s criteria is a really useful first step, and when someone comes to us already diagnosed this way, it gives us a clear, fast starting point to work from. In our clinical work, we then build on that picture by looking at what tipped the protective balance in the first place – because understanding the why is what keeps BV from coming back, and it works hand in hand with a solid bedside diagnosis rather than replacing it.
What to do next
If you have BV-type symptoms – a thin grey discharge, a fishy smell, irritation – it is worth being properly assessed rather than guessing, because the same symptoms overlap with several other conditions. A clinician can run Amsel’s criteria in the room and start you on the right track quickly.
If BV keeps coming back, or the picture is murky, consider going deeper with a comprehensive vaginal microbiome test so you can see exactly which bacteria are present, not just whether the bedside boxes are ticked.
For tailored help, you can ask Aunt Vadge’s Assistant – the chat widget in the bottom left of your screen – or book a consultation with one of our practitioners to work through what is driving your symptoms and how to address it.
Frequently asked questions
How many of the four criteria do I need for a BV diagnosis?
At least three of the four. Some clinicians use a modified version that diagnoses BV on two criteria, often clue cells plus a raised pH or characteristic discharge.1
Is Amsel’s criteria as good as the Nugent score?
For everyday diagnosis of symptomatic BV, yes – studies show the two agree well, and Amsel has the big advantage of being fast and cheap. Nugent remains the laboratory reference standard.37
Can I do Amsel’s criteria at home?
Not fully. You can test your vaginal pH at home, which covers one criterion, but the whiff test and clue cells need a clinician, KOH solution and a microscope.
Why might Amsel’s criteria miss my BV?
Its sensitivity is variable, so a real case can fall short of three positive criteria – especially if symptoms are mild, the sample is read by someone less experienced, or you are around menopause when pH naturally rises.1
Does a fishy smell on its own mean I have BV?
Not by itself. A fishy odour is a strong clue, but a positive whiff test specifically means the smell is released after KOH is added during an examination. A smell you notice at home is worth mentioning, but it is one piece of the picture and needs to be read alongside everything else.1
Should I get molecular testing instead?
For a straightforward, symptomatic episode, bedside Amsel’s criteria is usually enough. Molecular and sequencing tests come into their own for recurrent or confusing cases, where knowing the exact bacteria present helps guide treatment.8
How should I prepare for a BV test?
Try to avoid sex, douching and any vaginal washes, sprays or creams for about 24 hours before your appointment, as these can change your pH and affect the result.5
How long do BV test results take?
Bedside Amsel’s criteria can give an answer within the same appointment. A Nugent score or a molecular test is sent to a lab, so those usually take a few days.5
This article is general information and not a substitute for personalised medical advice. If you have symptoms or ongoing concerns, please see a qualified healthcare practitioner.
- Carlson K, Mikes BA. Amsel Criteria. StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; updated 2026 Jan 31.
- Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach D, Holmes KK. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med. 1983;74(1):14–22.
- Mohammadzadeh F, Dolatian M, Jorjani M, Alavi Majd H. Diagnostic Value of Amsel's Clinical Criteria for Diagnosis of Bacterial Vaginosis. Global Journal of Health Science. 2014;7(3).
- Mengistie Z, Woldeamanuel Y, Asrat D, Yigeremu M. Comparison of Clinical and Gram Stain Diagnosis Methods of Bacterial Vaginosis Among Pregnant Women in Ethiopia. JOURNAL OF CLINICAL AND DIAGNOSTIC RESEARCH. 2013.
- National Library of Medicine. Bacterial Vaginosis Test. MedlinePlus. Bethesda (MD): National Library of Medicine.
- Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. Journal of Clinical Microbiology. 1991;29(2):297–301.
- Challa A, Sood S, Kachhawa G, Upadhyay AD, Dwivedi SN, Gupta S. Diagnostic concordance between Amsel’s criteria and the Nugent scoring method in the assessment of bacterial vaginosis. Sex Health. 2022;18(6):512–514.
- Centers for Disease Control and Prevention. Bacterial Vaginosis – STI Treatment Guidelines, 2021. Atlanta (GA): CDC.
