COVID-19 is diagnosed primarily through two types of diagnostic tests: molecular tests (like PCR) that detect the virus’s genetic material, and antigen tests (the rapid tests you can take at home) that detect viral proteins. Both confirm an active infection, but they differ in speed, accuracy, and when they work best.
Molecular (PCR) Tests
Molecular tests, most commonly RT-PCR, are the gold standard for COVID-19 diagnosis. These tests detect fragments of the virus’s RNA and are typically processed in a lab, with results taking anywhere from a few hours to a couple of days. PCR tests have near-perfect specificity (around 99%), meaning a positive result almost certainly means you have the virus. False positives are rare and usually stem from sample contamination.
That said, PCR sensitivity is lower than many people assume. A large analysis published in the American Journal of Epidemiology estimated PCR sensitivity at about 68%, meaning roughly one in three infected people could get a negative result on any single test. This is partly because the test’s accuracy depends on timing, sample quality, and how much virus is present. PCR tests are still far more sensitive than rapid antigen tests, particularly early in infection or in people without symptoms.
Rapid Antigen Tests
Rapid antigen tests are the at-home kits most people are familiar with. They work by detecting proteins on the surface of the virus and deliver results in about 15 minutes. They’re convenient and widely available, but they’re less sensitive than PCR, especially in the first day or two of infection.
Antigen test sensitivity peaks around day four after symptoms begin, reaching about 77%. When compared against viral culture (which measures whether you’re carrying enough live virus to be contagious), sensitivity climbs to around 84%. This means antigen tests are better at identifying people who are likely infectious than they are at catching every single case.
If you’ve been exposed but don’t have symptoms, the FDA recommends waiting at least five full days after exposure before testing. Testing too early often produces false negatives because the virus hasn’t replicated enough to be detectable. For asymptomatic people, the recommended approach is serial testing: take a test, then test again 48 hours later, and again 48 hours after that, for a total of three tests over five days. Serial testing significantly improves your chances of catching an infection that a single test might miss.
When to Test After Exposure
Timing matters more than most people realize. After exposure, it typically takes two to five days (and sometimes longer) for the virus to build up enough in your body for any test to detect it. PCR tests can usually pick up an infection earlier than antigen tests because they amplify tiny amounts of genetic material, but even PCR can miss very early infections.
If you have symptoms, you can test right away with either type of test. If your first antigen test is negative but you still feel sick, test again in 48 hours. A single negative rapid test when you have cold or flu symptoms doesn’t rule out COVID.
Multiplex Tests for COVID, Flu, and RSV
Because COVID-19, influenza, and RSV can all cause similar symptoms, multiplex tests that screen for all three viruses at once are now available. Some of these are even sold as at-home kits. The CDC notes that molecular multiplex self-tests are preferred over antigen-based versions because of their higher sensitivity. These tests are especially useful during fall and winter when multiple respiratory viruses circulate at the same time and knowing which one you have can guide treatment decisions.
Antibody Tests Don’t Diagnose Active Infections
Antibody (serology) tests look for immune proteins in your blood that your body produced in response to a past infection or vaccination. They are not used to diagnose an active COVID-19 infection. Your body can take days to weeks to produce detectable antibodies after being infected, so these tests will often be negative during the period when you’re actually sick.
A positive antibody test can mean you were previously infected, that you’ve been vaccinated, or both. Not all antibody tests distinguish between vaccine-generated and infection-generated antibodies. Their main use is for population-level studies or for individuals trying to confirm a past infection.
How Variants Affect Test Accuracy
As SARS-CoV-2 mutates, some tests lose their ability to detect newer variants. This happens because diagnostic tests are designed to recognize specific genetic sequences or proteins, and mutations can alter those targets. Tests that rely on a single genetic target are most vulnerable. The FDA has revoked authorization for several tests that failed to reliably detect Omicron subvariants due to mutations in the nucleocapsid protein, which is the protein most antigen tests target.
Tests designed to detect multiple genetic targets are more resilient because even if one target mutates, the others can still flag the virus. Early data on Omicron indicated that antigen tests could still detect the variant but with reduced sensitivity. The FDA monitors authorized tests on an ongoing basis and publishes updates when specific products are affected by new mutations.
Imaging as a Diagnostic Tool
Chest imaging isn’t a standard way to diagnose COVID-19, but CT scans and chest X-rays play a supporting role in hospitalized patients. AI-assisted analysis of CT scans has achieved diagnostic accuracies between 90% and 96% for identifying COVID-19 pneumonia versus other lung conditions. In one study, radiologists reading CT scans with AI assistance were more accurate (90%) than those reading without it (85%). These tools help clinicians assess severity and guide treatment rather than replace standard viral testing.
Confirmed vs. Probable Cases
The CDC’s 2025 case definition draws a line between confirmed and probable COVID-19 cases. A confirmed case requires detection of the virus’s genetic material through a molecular test or genomic sequencing performed in a certified lab. A probable case is one where a lab-performed antigen test is positive but molecular confirmation hasn’t been done. In practice, this distinction matters more for public health tracking than for individual patients. If your at-home rapid test is positive and you have symptoms, you can be confident you have COVID-19 and act accordingly.