The most accurate COVID-19 test is the laboratory-based RT-PCR (reverse transcription polymerase chain reaction) test. PCR tests detect the virus’s genetic material and can pick up even tiny amounts of virus in a sample, making them the benchmark against which all other COVID-19 tests are measured. But accuracy depends heavily on timing, symptoms, and how you use the test, so the “best” test for your situation may not always be a PCR.
Why PCR Tests Are the Gold Standard
PCR tests work by amplifying traces of the virus’s genetic material until there’s enough to detect. This makes them extraordinarily sensitive. They can identify infections days before a rapid antigen test would turn positive, and they perform well regardless of whether you have symptoms. Every other type of COVID-19 test is evaluated by comparing its results to PCR.
That sensitivity comes with a tradeoff. PCR tests can remain positive for weeks after you’re no longer contagious, because they detect fragments of viral RNA that linger even after the live virus has cleared. So a positive PCR doesn’t always mean you’re currently infectious. It also means PCR isn’t ideal if you’re trying to figure out whether you’re still contagious after a known infection.
How Accurate Rapid Antigen Tests Really Are
Rapid antigen tests, the kind most people use at home, detect proteins on the surface of the virus rather than its genetic material. They require a higher viral load to trigger a positive result, which makes them less sensitive overall. In a CDC-supported study tracking daily testing from late 2022 through mid-2023, antigen tests had an overall sensitivity of just 47% compared to PCR. That means they missed roughly half of PCR-confirmed infections.
However, that number improves dramatically when symptoms are present. On days when people reported any COVID symptoms, sensitivity rose to 56% against PCR. When fever was present, it jumped to 77%. On days with no symptoms at all, sensitivity dropped to just 18%, meaning antigen tests missed more than four out of five infections in people who felt fine.
A separate university study found similar patterns: antigen tests reached 80% sensitivity in symptomatic people but only 41.2% in asymptomatic individuals. The takeaway is that a negative rapid test is far more reliable when you have symptoms than when you don’t.
Serial Testing Closes the Gap
A single rapid antigen test is a snapshot. Testing multiple times over several days significantly improves accuracy. Research from Northwestern University found that symptomatic patients who took two antigen tests 48 hours apart reached a cumulative sensitivity of 93.4% compared to PCR. That’s close to PCR-level performance using cheap, at-home tests.
For asymptomatic people, two tests 48 hours apart reached 62.7% sensitivity, which then improved to 79% with a third test at 48 hours. The FDA recommends this three-test approach if you’ve been exposed but don’t have symptoms: wait at least five full days after exposure, then test three times with 48 hours between each test. Testing too early, before the virus has had time to replicate, is one of the most common reasons for a false negative.
At-Home Molecular Tests: A Middle Ground
A newer category of at-home tests uses a molecular technique called LAMP (loop-mediated isothermal amplification), which detects viral genetic material like PCR but in a simpler, portable format. These tests sit between rapid antigen tests and full laboratory PCR in terms of accuracy.
In a head-to-head comparison published in Frontiers in Microbiology, the LAMP-based test achieved 70.9% sensitivity compared to PCR, while the rapid antigen test in the same study reached only 54.6%. The difference was especially pronounced in people with lower viral loads: among samples with moderate virus levels, the LAMP test detected 87.5% of infections versus 60% for the antigen test. In the early phase of illness (within seven days of symptom onset), LAMP detected 86.8% of infections compared to 72.4% for the antigen test.
When researchers extracted and purified the RNA before running the LAMP test (a step done in a lab, not at home), sensitivity climbed to 92.9%. The at-home version won’t match that, but it still outperforms standard lateral flow antigen strips.
Timing Matters More Than Test Type
Even the most accurate test will miss an infection if you take it at the wrong time. After exposure, the virus typically needs two to five days (sometimes longer) to replicate enough to be detectable. PCR tests can pick up the virus slightly earlier in this window than antigen tests, but neither is reliable on day one or two after exposure.
For antigen tests, the sweet spot is when viral load peaks, which generally coincides with the first few days of symptoms. If you develop a sore throat or congestion and test negative, repeating the test 48 hours later is more informative than assuming you’re in the clear. The FDA specifically notes that at-home antigen tests are less likely to detect the virus early in an infection or in people without symptoms.
How Variants Affect Test Performance
New variants can change the proteins and genetic sequences that tests are designed to detect. The FDA monitors this continuously and has flagged specific tests whose performance dropped with certain Omicron subvariants. Early data showed that antigen tests could still detect Omicron but with reduced sensitivity in some cases. A small number of affected antigen tests had their emergency authorizations revoked.
Most molecular tests are designed to detect multiple genetic targets at once, which provides a buffer. If a mutation disrupts one target, the other targets still work. For example, some PCR tests experienced “gene dropout” where one of their targets failed against certain variants, but because they check two or three targets simultaneously, overall sensitivity remained intact. A few molecular tests that relied too heavily on a single target did see meaningful drops in accuracy with XBB and BA.2.75 subvariants.
Choosing the Right Test for Your Situation
If you need the highest possible accuracy for a single test, a laboratory PCR is your best option. It’s the right choice when you need a definitive answer: before a medical procedure, after a known high-risk exposure, or when symptoms are ambiguous.
If you want a quick answer and you currently have symptoms, a rapid antigen test is reasonable, especially if you’re willing to repeat it 48 hours later. Two antigen tests taken 48 hours apart while symptomatic approach 93% sensitivity, making serial testing a practical and accessible strategy.
If you’re asymptomatic and trying to confirm you’re not infected (before visiting someone vulnerable, for example), a single rapid antigen test is the least reliable approach. Three tests over four days, or a single PCR test taken at least five days after exposure, will give you a much clearer picture. At-home molecular tests based on LAMP technology offer a step up in accuracy over antigen tests if you can find them, though they’re less widely available and more expensive.