The whiff test, also called the KOH amine test, is a simple bedside laboratory technique used during clinical evaluation of vaginal discharge. The clinician collects a small sample of vaginal fluid on a glass slide, adds a drop of 10 percent potassium hydroxide solution, and immediately smells the slide. The release of a strong fishy or amine-like odour after the KOH is added indicates a positive result. The test is rapid, costs almost nothing, and contributes one of the four Amsel criteria used to diagnose bacterial vaginosis (BV) in routine outpatient care.
Behind the simplicity is a useful biochemistry. Bacterial vaginosis develops when the normal lactobacillus-dominant vaginal flora is replaced by anaerobic species such as Gardnerella vaginalis, Mobiluncus and various Prevotella organisms. These anaerobes produce volatile amines — putrescine, cadaverine and trimethylamine — that remain in the vaginal fluid in a non-volatile, salt-bound form. Adding the alkaline potassium hydroxide drives off the protons and releases the amines as gas, producing the characteristic odour. A normal lactobacillus-dominant flora does not generate these amines, so a healthy sample produces no odour when KOH is added.
The test has been part of standard outpatient gynaecological practice since the early 1980s and has survived the introduction of more sophisticated molecular alternatives because of its sheer practicality. A clinician can complete the entire procedure in less than thirty seconds using equipment already on the speculum tray, which makes it ideal for high-volume settings such as primary care offices, sexually transmitted infection clinics and emergency departments where time and laboratory resources are limited.
Reagent: 10% potassium hydroxide. Procedure: drop KOH onto vaginal fluid sample, smell immediately. Positive result: strong fishy or amine odour. Place in workup: one of four Amsel criteria for bacterial vaginosis. Sensitivity ~67–70%, specificity ~94%. Performed at bedside or in clinic; result available in seconds. Negative test does not rule out BV.
The Amsel criteria are the standard clinical framework used to diagnose bacterial vaginosis without sending samples to a reference laboratory. A patient meets the diagnosis if at least three of the four criteria are present. The first criterion is a homogeneous, thin, white-grey vaginal discharge that coats the vaginal walls.
The second is a vaginal pH greater than 4.5, measured with a strip of pH paper applied to a swab of vaginal fluid. The third is a positive whiff test, which is what this article focuses on. The fourth is the presence of clue cells on a saline wet mount viewed under a microscope.
The criteria were proposed in 1983 by Amsel and colleagues and have remained the foundation of clinical BV diagnosis ever since because they perform reasonably well at the bedside without requiring specialised laboratory infrastructure. None of the four criteria alone is sensitive enough to diagnose BV, but the combination produces a sensitivity of around 90 percent compared with the gold-standard Gram stain Nugent score. The whiff test contributes a quick, intuitive yes-or-no signal that integrates well with the rest of the bedside examination.
One nuance worth understanding is that not every clinician applies the Amsel criteria the same way. Some practitioners weight the four findings equally, while others place more diagnostic weight on the presence of clue cells because that single finding has the highest specificity for BV. National guidelines, including those from the CDC and the Royal College of Obstetricians and Gynaecologists, recommend the three-of-four threshold but acknowledge that experienced clinicians may diagnose BV based on a strong constellation of two findings combined with classic patient history.
The four criteria are also not equally easy to apply in every clinical setting. Microscopy for clue cells requires a microscope and time, which busy walk-in clinics may not have. The whiff test and pH measurement are achievable anywhere, while the discharge appearance is a subjective observation that can vary between examiners. The complementary nature of the four criteria explains why the Amsel framework has remained durable despite its imperfections.
White or grey, thin and uniformly distributed across the vaginal walls. Often noticed by the patient as increased discharge with a noticeable odour, particularly after intercourse.
Measured with pH paper applied to a vaginal swab away from cervical mucus. Healthy lactobacillus-dominated flora produces a pH of 3.8–4.2. BV elevates pH because lactobacilli are reduced or absent.
Drop 10% KOH on a sample and smell immediately. Fishy odour signals release of volatile amines from anaerobic metabolism. Negative whiff test does not rule out BV but lowers probability.
Vaginal epithelial cells coated with adherent bacteria visible under microscopy. Edges of the cell appear stippled rather than crisp. Considered the most specific of the four criteria.
The procedure takes less than a minute when integrated into a standard speculum exam. A sterile cotton swab is rolled across the lateral vaginal wall to collect a sample of discharge. The swab is then either rolled onto a glass slide for the wet mount portion or placed in saline for later examination.
For the whiff test specifically, a small amount of the fluid is placed on a clean slide. A drop of 10 percent potassium hydroxide solution is added on top of the fluid. The slide is held a few centimetres from the clinician's nose, and the odour is assessed within seconds, before the volatile amines disperse.
Some clinics combine the whiff step with the wet mount examination to save time. The KOH preparation also serves the secondary purpose of dissolving epithelial cells and revealing fungal hyphae for candidiasis evaluation. A clinician evaluating mixed vaginitis often uses the same KOH slide first to whiff for amines and then under the microscope to look for hyphae. The economy of using one preparation for two clinical questions is a small but real benefit in busy outpatient settings.
The KOH solution itself deserves a moment of attention. The 10 percent strength is calibrated to dissolve epithelial cell debris while reliably releasing amines without burning the clinician's nose. Higher concentrations release amines faster but produce more caustic fumes; lower concentrations may not fully drive the amine release. Pre-mixed clinical KOH from medical supply vendors is the safest source because home-mixed solutions are subject to evaporation and concentration drift over time.
Sterile cotton swabs, glass microscope slides, 10 percent KOH solution in a dropper bottle, saline for the wet mount comparison, pH paper for the parallel pH measurement, and basic personal protective equipment. The total reagent cost per test is well under a dollar, which is why the test remains widely used despite the availability of more sensitive molecular alternatives.
The test is most accurate when performed immediately after sample collection. Volatile amines disperse rapidly once released, so smelling the slide within five to ten seconds of adding KOH is essential. Delayed assessment produces false negatives even when the underlying flora would have produced a positive result.
A strong, distinctly fishy odour is reported as positive. A faint or ambiguous odour should be recorded as negative because false positives can lead to overtreatment. The result is binary in clinical documentation, although clinicians often note the strength of the odour informally.
Recent intercourse can produce a transient fishy odour from semen exposure even without BV, because seminal fluid is alkaline. Menstrual blood can interfere with the assessment. Both conditions should be noted and the test deferred or interpreted with caution. KOH dropper bottles that have been opened for many months may lose potency.
The result is documented in the clinical note alongside the other Amsel criteria. Most electronic health records have structured fields for vaginal pH, whiff test result, clue cells and discharge appearance. Capturing all four together enables retrospective review of diagnostic accuracy in audit settings.
The test is skipped during heavy menstrual bleeding, soon after intercourse, or when a molecular BV panel has already been sent. Some clinicians also skip it in patients with strong olfactory aversion or suspected anosmia in the examiner. Always document why a criterion was not assessed.
The whiff test alone is not diagnostic of bacterial vaginosis. Published sensitivity ranges from about 67 to 70 percent, meaning that one in three patients with BV will not produce a positive whiff. Specificity is much higher at around 94 percent, meaning that a positive whiff test in a symptomatic patient strongly raises the probability of BV. The asymmetric performance shapes how the test should be used: a positive whiff is meaningful, while a negative whiff cannot exclude the diagnosis on its own.
Accuracy also depends on the examiner's olfactory ability, which is rarely formally assessed. Studies of medical trainees have shown surprising variability in odour detection thresholds, particularly among older clinicians and those with chronic rhinosinusitis. Some institutions have moved toward objective alternatives — molecular probe panels, DNA-based BV testing kits and the OSOM BVBlue cassette — which remove the human variable from the assessment. The whiff test remains useful precisely because it is fast, cheap and integrates with the existing speculum exam, but its limits should be understood by every clinician who relies on it.
Inter-rater reliability is another consideration. Two clinicians smelling the same slide do not always agree on whether the result is positive, especially when the odour is faint. Studies of paired examiners have shown agreement rates around 85 percent for clear positives and far lower for borderline cases. This variability is one reason large clinical trials of BV treatments now use molecular endpoints rather than the whiff test or even Amsel criteria as their primary diagnostic measure.
A positive whiff test is most often associated with bacterial vaginosis, but the differential is worth knowing. Trichomonas vaginalis infection produces volatile amines through similar anaerobic metabolic pathways and frequently produces a positive whiff. Patients with trichomoniasis typically also report frothy yellow-green discharge, vaginal soreness and dyspareunia, and the protozoan can sometimes be seen swimming on the saline wet mount. Coinfection with both BV and trichomoniasis is common, which is one reason guidelines recommend testing for trichomonas via NAAT in any patient with persistent or recurrent BV-like symptoms.
Candidiasis is the third common cause of vaginitis and typically produces a negative whiff. The discharge is white, thick and often described as cottage-cheese-like. Vaginal pH stays normal at 3.8 to 4.2, distinguishing candidiasis from BV. Mixed vaginitis with both BV and candidiasis is reasonably common, particularly in patients on recent antibiotics, and the combination produces both a positive whiff and visible hyphae on KOH preparation. Recognising mixed pictures avoids the trap of treating only one of the two conditions and missing the other.
Atrophic vaginitis in postmenopausal women can sometimes produce mild discharge changes that complicate BV diagnosis. The vaginal pH may rise above 4.5 due to oestrogen deficiency rather than anaerobic overgrowth, which makes the pH criterion less specific in this population. Whiff tests in atrophic vaginitis are usually negative because the underlying microbiome shift is different from that of true BV. Considering the patient's age and hormonal status when interpreting the whiff result avoids overdiagnosis of BV in postmenopausal patients.
Several commercially available molecular tests now compete with the Amsel criteria for BV diagnosis. The OSOM BVBlue is a chromogenic point-of-care cassette that detects sialidase activity from anaerobic bacteria. The BD MAX Vaginal Panel and Aptima BV are NAAT-based assays that detect specific bacterial species and quantify the relative abundance of lactobacillus and anaerobes. These molecular tests have higher sensitivity than the Amsel criteria, particularly in patients with subtle or atypical presentations, and are increasingly preferred when the local laboratory infrastructure supports them.
The Nugent score remains the research gold standard. It assigns numeric values to lactobacillus, Gardnerella and Mobiluncus morphotypes on a Gram-stained vaginal smear and produces a composite score from 0 to 10. Scores of 7 or higher are diagnostic of BV; scores of 4 to 6 are intermediate; scores of 0 to 3 are normal. Most epidemiological BV studies use the Nugent score because it is reproducible and not subject to the patient-clinician interaction of the Amsel approach.
Cost is the practical barrier to widespread molecular testing. NAAT-based BV panels can run $80 to $200 per test depending on the laboratory contract, compared with pennies for the whiff test and Amsel criteria. Many health systems use the bedside approach as a screening step and reserve molecular testing for recurrent or refractory cases. Insurance coverage policies also vary, with some payers requiring documented Amsel criteria findings before approving molecular BV panels.
Bacterial vaginosis is the most common cause of abnormal vaginal discharge in women of reproductive age. While many patients are asymptomatic, untreated BV is associated with several adverse health outcomes that elevate it from a comfort-and-quality-of-life concern to a clinically significant infection. Pregnant women with BV face increased risk of preterm birth, late miscarriage, premature rupture of membranes, and post-caesarean endometritis. Non-pregnant women with BV face increased risk of pelvic inflammatory disease following gynaecological procedures, surgical site infection after hysterectomy, and a documented increase in HIV acquisition risk during exposure events.
BV also increases susceptibility to other sexually transmitted infections, including chlamydia, gonorrhoea and herpes simplex virus. The disrupted vaginal microbiome reduces local antimicrobial peptide production and creates conditions favourable to pathogen establishment. These public-health considerations argue for accurate diagnosis and treatment even in mildly symptomatic or asymptomatic high-risk patients. The whiff test contributes meaningfully to that diagnostic process even though it cannot stand alone.
Recurrent BV is also a substantial clinical problem, with up to 30 percent of patients experiencing relapse within three months of treatment and over 50 percent within twelve months. The drivers of recurrence are still incompletely understood and likely involve persistent biofilm formation by Gardnerella vaginalis on the vaginal epithelium. The whiff test continues to be useful in following up patients with suspected recurrence because it provides a quick yes-or-no signal that can guide whether to pursue further microbiological work-up.
500 mg twice daily for 7 days. First-line treatment endorsed by CDC and most professional society guidelines. Avoid alcohol during and for 24 hours after treatment due to disulfiram-like reaction.
0.75% gel administered intravaginally once daily for 5 days. Equivalent efficacy to oral therapy with fewer systemic side effects. Often preferred for patients who do not tolerate oral metronidazole.
2% cream administered intravaginally at bedtime for 7 days. Alternative for patients with metronidazole allergy or intolerance. May weaken latex condoms during and after use.
2 g once daily for 2 days, or 1 g daily for 5 days. Alternative oral nitroimidazole with similar efficacy and longer half-life than metronidazole. Same alcohol caution applies.
Patients with frequent recurrence may benefit from extended suppressive courses such as twice-weekly metronidazole gel for several months after initial treatment. New strategies include vaginal lactobacillus replacement therapy.
Routine treatment of sexual partners is not recommended for BV. The condition is not classified as a sexually transmitted infection, although sexual activity influences microbiome composition.
The most common error is delayed assessment. Volatile amines escape the slide within seconds, and a clinician who walks across the room before smelling the slide will miss positive results. Performing the test at the bedside immediately after KOH addition is essential. The second common error is using diluted or expired KOH solution. Bottles that have sat open for more than six months can lose potency, particularly in clinics with poor storage discipline. Replacing dropper bottles every six months and storing them upright with caps tight is the simplest fix.
The third common error is testing during inappropriate windows. Recent intercourse leaves alkaline seminal fluid in the vagina, which can mimic the BV amine signal. Menstrual blood interferes with both the whiff test and the wet mount. Patients should ideally be tested at least 24 to 48 hours after the last intercourse and outside the active menstrual period. When timing is not ideal, the test result should still be recorded but interpreted with caution alongside the other clinical findings.
One subtle but important issue is the ergonomics of bedside testing. Clinicians who lean too close to the slide can become desensitised to subsequent samples through olfactory adaptation, particularly during high-volume STI clinics where multiple BV evaluations occur back to back. Pausing between patients, stepping out of the exam room briefly, and rotating responsibility for the whiff component among team members are all simple practices that maintain assessment accuracy across a busy session.
Trainee education on the whiff test has shifted toward simulation and structured supervision. Many residency programmes now incorporate paired observation with attending physicians during early clinical exposure so that learners can calibrate their own olfactory sensitivity against an experienced examiner. Building this baseline early avoids the silent skill drift that can otherwise leave clinicians overly confident in a finding that depends on their own perception.
Documenting variability between examiners during teaching rounds reinforces the message that the test is operator-dependent.