COVID tests are generally reliable, but their accuracy varies significantly depending on the type of test, whether you have symptoms, and when you test relative to your exposure. PCR tests are the most accurate, with specificity near 100% and peak sensitivity around 93%. Rapid antigen tests (the at-home kind) are less sensitive, catching about 80% of infections in people with symptoms but only around 40% in people without symptoms on a single test.
Understanding what drives these numbers helps you interpret your own results and know when to trust a negative.
PCR Tests: The Most Reliable Option
PCR tests, the kind processed in a lab from a nasal or throat swab, remain the gold standard. Their specificity is close to 100%, meaning false positives are extremely rare. When a PCR test says you’re positive, you almost certainly are.
Sensitivity (the ability to catch a real infection) depends heavily on timing. On the day of infection, a PCR test has essentially 0% sensitivity because the virus hasn’t replicated enough to detect. By three days after infection, sensitivity climbs to about 48%. It peaks at roughly 93% around four to five days after infection and stays above 88% through the second week. By the day symptoms first appear, sensitivity is around 86%.
The FDA generally required at least 95% sensitivity and specificity for molecular tests to receive emergency authorization during the pandemic. Most authorized PCR tests meet or exceed that threshold under ideal conditions.
Rapid Antigen Tests: Convenient but Less Sensitive
Rapid antigen tests, including the at-home kits most people are familiar with, trade some accuracy for speed and convenience. They detect viral proteins rather than genetic material, and they need a higher amount of virus to trigger a positive result. Most antigen tests reliably detect infections when viral loads reach about 3 million RNA copies per milliliter, a threshold that corresponds to a moderate to high viral load. PCR tests can detect far lower concentrations.
In studies comparing antigen tests to PCR, the overall sensitivity of antigen tests is around 47% when using PCR as the reference. That number sounds alarming, but it’s partly because PCR picks up very low-level infections, including people who are no longer contagious. When measured against viral culture (which indicates whether someone is actually shedding live, infectious virus), antigen test sensitivity rises to about 80%. In other words, rapid tests are better at identifying people who are contagious than at catching every trace of infection.
The FDA set a lower bar for antigen tests: at least 80% sensitivity and 95% specificity for authorization, with even lower sensitivity accepted for at-home tests as long as instructions included guidance on repeat testing.
Symptoms Make a Big Difference
Whether or not you have symptoms dramatically affects how well a rapid test performs. In symptomatic people, a single rapid antigen test catches about 80% of infections. In asymptomatic people, that drops to roughly 41%. The reason is straightforward: symptoms typically appear when viral load is high, which is exactly when antigen tests work best. Before symptoms develop, or if you never develop them, the virus may be present at levels too low for an antigen test to detect.
This is why a negative rapid test when you feel fine doesn’t guarantee you’re not infected. It’s also why public health guidance has consistently recommended repeat testing.
Serial Testing Closes the Gap
Testing more than once significantly improves rapid antigen test accuracy. A large NIH study tracked what happens when people test every 48 hours:
- Symptomatic people: Sensitivity rose from 83% on the first test to 93% on the second test two days later.
- Asymptomatic people: Sensitivity jumped from 39% on the first test to 63% on the second, and 79% by the third test.
Two tests spaced 48 hours apart bring rapid antigen accuracy in symptomatic people close to PCR levels. For asymptomatic screening, three tests over four to five days are needed to reach a similar level of confidence. If you’re testing after a known exposure but feel fine, a single negative test on day one is not especially reassuring. Testing again on day three and day five gives you a much more reliable picture.
Timing Relative to Exposure
The single biggest factor in test accuracy is when you test. The virus needs two to five days after exposure (sometimes longer) to replicate enough for any test to detect it. Testing the day after a known exposure is almost guaranteed to return a negative result regardless of whether you’re infected.
Viral load follows a predictable curve: it starts low, rises quickly, peaks around the time symptoms begin, and gradually declines as the immune system responds. Both PCR and antigen tests perform best during the rising and peak phases. PCR holds its accuracy longer into the declining phase because it can detect smaller amounts of viral material. Antigen tests tend to turn negative sooner, sometimes while PCR is still positive but when the person is likely no longer very contagious.
False Positives Are Rare but Real
Specificity for both PCR and antigen tests is high, meaning false positives are uncommon. For rapid antigen tests, specificity runs between 98% and 99%. But “uncommon” is not zero.
A large analysis from the NIH’s RADx program, covering more than 11,000 participants who took serial antigen tests alongside confirmatory PCR, found that 1.7% of participants had at least one false positive on a rapid antigen test. Most of these were one-time events. However, a small group (about 7% of those with false positives) had persistent false positives, meaning the antigen test kept showing a faint line even though PCR repeatedly confirmed no infection.
Nearly all the persistent false-positive cases were women, and most reported having an autoimmune condition. Researchers suspect a cross-reaction between the test and immune proteins like rheumatoid factor, which the immune system produces in certain autoimmune diseases. If you consistently get faint positive lines on rapid tests but feel fine and have no known exposure, confirming with a PCR test is the clearest way to resolve the ambiguity.
Omicron Variants and Newer Strains
As the virus has evolved, concerns about test accuracy have resurfaced with each new variant. CDC data from late 2022 through mid-2023, during a period of circulating Omicron subvariants and widespread population immunity, found that antigen test performance remained largely consistent with earlier pandemic findings. Sensitivity was 47% against PCR and 80% against culture, numbers very similar to earlier studies with different variants.
The stability of these numbers across different variants suggests that current rapid tests are not significantly less accurate against newer strains than they were against earlier ones. Differences in reported sensitivity between studies are more likely explained by variations in participant immunity, sampling technique, and the sensitivity of the PCR test used as a reference than by the variant itself.
What a Negative Test Actually Means
A negative PCR test taken four or more days after exposure is highly reliable. A negative rapid antigen test is a snapshot of that moment: it means your viral load was below the test’s detection threshold at the time you swabbed. It does not rule out early infection, and it does not mean you won’t test positive tomorrow.
If you’re symptomatic and test negative on a rapid test, testing again in 48 hours catches most of the infections the first test missed. If you’re asymptomatic but had a known exposure, plan on testing at least twice, ideally three times over five days. A positive rapid test, on the other hand, is almost always correct. The specificity of these tests is high enough that a clear positive line reliably indicates infection.