Yellow fever is diagnosed through a combination of clinical evaluation and laboratory testing, with the specific test depending on how many days have passed since symptoms began. In the first ten days, doctors can detect the virus itself in a blood sample using molecular testing. After about a week, antibody-based blood tests become the primary tool. Because yellow fever shares symptoms with many other tropical infections, laboratory confirmation is always required.
Clinical Suspicion: The Starting Point
The World Health Organization defines a suspected case as any person with sudden onset of fever plus jaundice appearing within 14 days of the first symptoms. That combination, especially in someone who has traveled to or lives in a region where yellow fever circulates (parts of sub-Saharan Africa and South America), is the trigger for laboratory testing.
Early yellow fever looks like many other tropical illnesses: fever, headache, muscle pain, nausea. It only becomes more distinctive when jaundice, bleeding, or organ damage develops. This overlap with malaria, dengue, viral hepatitis, and hemorrhagic fevers like Rift Valley fever means that clinical signs alone are never enough to confirm the diagnosis. In one study from the Democratic Republic of Congo, nearly 44% of suspected yellow fever cases that tested negative turned out to have a hepatitis virus instead, most commonly hepatitis B.
Molecular Testing in the First 10 Days
The most direct way to identify yellow fever is a test called RT-PCR, which detects the virus’s genetic material in a blood or serum sample. This is considered the gold standard for early diagnosis and works within days 1 through 10 after symptoms start. After that window closes, viral levels in the blood typically drop below what the test can pick up.
Timing matters. If you develop symptoms and seek care within the first week or so, a PCR test can confirm yellow fever quickly and specifically. The sample needs to reach a qualified laboratory while still viable: serum should be shipped on ice within 48 hours of collection, or stored at refrigerator temperature (4 to 8°C) until transport is arranged. If serum can’t be separated from whole blood right away, the whole blood sample must be kept refrigerated (never frozen) and sent to the lab within 24 hours.
Urine is also being explored as a sample type for molecular diagnosis, since viral RNA may be detectable in urine even during the recovery phase when it has already cleared from the blood.
Antibody Testing After Day 7
Once the virus clears from the bloodstream, diagnosis shifts to detecting the immune system’s response. The primary tool is an IgM antibody test, typically performed with a method called ELISA. IgM antibodies are the first wave your body produces against a new infection, and they become detectable around 7 to 10 days after symptoms begin. At that point, the test is about 95% sensitive.
A single positive IgM result during the late acute or early recovery phase is enough for a presumptive diagnosis. For a fully confirmed case, the WHO criteria require one of the following: a positive IgM result, a fourfold rise in antibody levels between two blood samples taken days apart, or detection of yellow fever-specific neutralizing antibodies.
Why Confirmation Gets Complicated
Yellow fever belongs to a family of viruses called flaviviruses, which also includes dengue, Zika, and West Nile. These viruses are structurally similar enough that antibody tests frequently cross-react. A person who had dengue last year, or who received the yellow fever vaccine, can produce antibodies that light up on an ELISA test even without a current yellow fever infection. In regions where multiple flaviviruses circulate (which includes most areas where yellow fever is found), this cross-reactivity is a serious diagnostic challenge.
This is where a more specialized test comes in: the plaque reduction neutralization test, or PRNT. It measures antibodies that specifically neutralize the yellow fever virus, making it far more precise than a standard ELISA. The PRNT is considered the gold standard for distinguishing yellow fever from other flavivirus infections. A well-validated PRNT can correctly identify yellow fever antibodies even in samples spiked with antibodies against dengue, Zika, or Japanese encephalitis, showing no meaningful cross-reactivity.
The tradeoff is that the PRNT is slower, more labor-intensive, and only available at specialized reference laboratories. It is typically reserved for confirming ambiguous results rather than used as a first-line test.
The Vaccine Complication
The yellow fever vaccine uses a live, weakened form of the virus, which means it triggers the same types of antibodies as a natural infection. If you received the vaccine within the past 30 days and then develop a fever, standard antibody tests cannot reliably tell whether your symptoms are from the vaccine strain or a wild infection. The WHO explicitly flags this as a diagnostic gap: ELISA testing alone cannot confirm a current yellow fever infection when the antibody response could be from past infection, another flavivirus, or vaccination.
For this reason, the WHO’s confirmed case definition requires that the person had no yellow fever vaccination within 30 days before symptom onset (for antibody-based confirmation) or within 14 days (for PCR-based confirmation). When recent vaccination is in the picture, molecular testing or antigen detection becomes especially important, since newer diagnostic assays are being designed to distinguish vaccine strains from wild-type virus.
What Happens With Your Sample
A minimum of 0.5 milliliters of separated serum is needed for confirmatory testing at a reference laboratory. In many countries, initial testing happens at a national laboratory, and samples that need further confirmation are shipped to WHO-designated regional reference labs. The cold chain is critical: samples travel on ice packs at 2 to 8°C, which keeps them stable for up to three days in a well-insulated container. For longer transit times, dry ice at minus 80°C is recommended.
The testing algorithm generally follows a stepped approach. Molecular testing (PCR) is attempted first if the sample was collected early enough. If the sample arrives after day 10, or if PCR is negative but clinical suspicion remains, IgM ELISA is performed. Positive or ambiguous ELISA results are then sent for PRNT confirmation.
Conditions That Mimic Yellow Fever
Because so many tropical diseases cause fever and jaundice, diagnostic testing for yellow fever often runs alongside tests for other infections. Malaria is one of the most important to rule out, partly because it is so common in overlapping regions and partly because coinfection with both malaria and yellow fever has been documented. Dengue can cause similar fever, headache, and muscle pain, though it less commonly damages the liver to the degree yellow fever does. Hepatitis A through E can all present with jaundice and liver inflammation. Hemorrhagic fever viruses, including Ebola, Lassa, and Crimean-Congo hemorrhagic fever, can mimic the severe bleeding and organ failure seen in advanced yellow fever cases.
A broad initial workup is standard. Malaria rapid tests or blood smears, hepatitis panels, and dengue testing are often run in parallel, especially since a negative result for one pathogen does not rule out another.
Tissue Diagnosis After Death
In fatal cases where blood-based testing was not performed in time, yellow fever can be diagnosed by examining liver tissue. The hallmark finding is a specific pattern of liver cell death concentrated in the middle zone of the liver lobule, with relatively little inflammatory response. Damaged liver cells form distinctive round or oval structures historically called Councilman bodies, which contain fat droplets and a pigment called ceroid. The amount of ceroid pigment correlates with the severity of liver damage. These features, taken together, are characteristic enough to distinguish yellow fever from other causes of liver failure, though this type of diagnosis is primarily relevant for surveillance and outbreak investigation rather than patient care.