How Long Is COVID Contagious? What the CDC Says

Most people with COVID-19 are contagious for roughly five to ten days after symptoms begin, with the highest risk of spreading the virus concentrated around days two and three. The CDC’s current guidance, updated in March 2024, moved away from a fixed isolation clock and instead ties your return to normal activities to how you feel: fever-free for at least 24 hours without medication, and symptoms improving. That shift reflects years of accumulated evidence about how viral shedding actually works, but the biology is messier than any single rule can capture.

What the CDC Now Recommends

For most of the pandemic, the CDC anchored its isolation guidance to a specific number of days. The original recommendation was ten days from symptom onset. Modeling studies supported that timeline: one analysis estimated the risk of releasing someone still infectious after a fixed ten-day isolation was low, between zero and about seven percent, though people were spending roughly five to eight unnecessary days in isolation beyond when they were actually contagious.¹ In late December 2021, the CDC shortened the recommended isolation to five days for people whose symptoms were resolving, followed by five days of masking. That update drew immediate scrutiny from researchers who noted that many people still tested positive on rapid antigen tests after just five days of isolation.

By March 2024, the CDC folded COVID into a broader respiratory virus framework alongside flu and RSV. The new approach no longer specifies a day count. Instead, you stay home while you have a fever or feel acutely ill, and you can return to regular activities once your fever has been gone for a full 24 hours without antipyretics and your overall symptoms are trending better. The CDC still recommends taking extra precautions for the following five days, like wearing a mask in indoor public spaces and improving ventilation. This symptom-based standard acknowledges what the research consistently shows: the infectious window varies from person to person, and symptoms are a better proxy for contagiousness than a calendar.

When Viral Shedding Peaks

The amount of virus in your respiratory tract rises fast once an infection takes hold. Data from recent surveillance cohorts show that viral loads tend to peak around the second day of symptoms, though people who were vaccinated within the past year sometimes see peak loads shifted to the third day.² The onset of contagiousness tends to coincide with a sharp increase in upper respiratory symptoms. A community cohort study found that symptoms like a runny nose, sore throat, and hoarse voice showed the steepest rise on the first day that live virus could be cultured from participants.³

The good news is that infectious virus disappears from most people well before their PCR tests turn negative. A systematic review of viral culture studies found that researchers could not isolate live virus from samples taken after day eight of illness, even when patients still carried high levels of viral RNA.⁴ Contact-tracing data point in the same direction. One study concluded that the period of highest transmissibility stretches from about four days before symptom onset to five days after, and that no onward transmission occurred when close contacts were first exposed more than five days after the infected person’s symptoms appeared.⁵

Why You Can Test Positive Long After You Stop Being Contagious

One of the most persistent sources of confusion throughout the pandemic has been the gap between a positive test and actual infectiousness. PCR tests detect genetic fragments of the virus, and those fragments can linger in your airways for weeks after the live virus is gone. The systematic review mentioned above found that six out of eight studies documented RNA shedding lasting more than 14 days, far longer than the window in which live virus could be grown in a lab.⁴ A review in Nature Reviews Microbiology put it plainly: RNA can be picked up by PCR long after infectious virus is undetectable, and rapid antigen tests are a better proxy for whether someone is still capable of spreading the infection.⁶

Some people who have recovered also experience a re-positive PCR result days or weeks later. Research suggests these re-positive results reflect the shedding of non-infectious viral RNA fragments rather than a return of active infection. One study found that patients who re-tested positive showed no corresponding rise in inflammatory markers or antibody fluctuations, further supporting the idea that the viral debris picked up by the test was not capable of causing disease or spreading to others.⁷

The Invisible Spread Before Symptoms Start

A large fraction of COVID transmission happens before the infected person even knows they are sick. One modeling study that accounted for variation in when people become most infectious estimated that roughly 65 percent of transmissions occurred before symptom onset.⁸ A separate decision-analytic model published in JAMA Network Open arrived at a broadly consistent picture: under baseline assumptions, about 59 percent of all transmission came from people without symptoms at the time they spread the virus. Of that, roughly 35 percent came from presymptomatic individuals who would later develop symptoms, and about 24 percent from people who never developed symptoms at all.⁹

People who remain truly asymptomatic throughout their infection are less likely to spread the virus than those who are presymptomatic or overtly sick. A meta-analysis found the transmission rate from asymptomatic index cases was about 1.8 per 100 person-days, compared with roughly 5.0 from presymptomatic cases and 5.3 from symptomatic ones.¹⁰ So while truly asymptomatic infections are less contagious on a per-contact basis, the sheer volume of presymptomatic transmission is what has made COVID so difficult to contain with symptom-based screening alone.

How Vaccination Changes the Contagious Window

Vaccination does not prevent infection reliably, but it does appear to shorten the period during which you shed live virus. A longitudinal cohort study of Delta variant infections found that among people shedding infectious virus, the median duration was reduced from about seven and a half days in unvaccinated participants to six days in vaccinated ones. From day six onward after symptom onset, vaccinated individuals had roughly 58 percent lower odds of still shedding infectious virus compared to unvaccinated individuals.¹¹ Another study was more dramatic in its findings: fully vaccinated individuals shed viable virus for a median of four days after symptom onset, compared with eight days for partially vaccinated and ten days for unvaccinated people.¹²

The downstream effect on household transmission is substantial. A meta-analysis of household secondary attack rates found that unvaccinated index cases spread the virus to about 30 percent of their household contacts, while fully vaccinated index cases transmitted to about 14 percent. The difference was statistically clear. On the receiving end, unvaccinated contacts had a roughly 34 percent chance of catching the virus from an infected household member, versus about 14 percent for fully vaccinated contacts.¹³ Vaccination benefits both the infected person’s contagious timeline and the vulnerability of the people around them.

How Variants Changed the Timeline

The shift from Delta to Omicron brought a noticeable change in how quickly symptoms appear after exposure. A French nationwide study estimated that the mean incubation period fell from about 4.4 days with Delta to roughly 3.6 days with Omicron, compared with 5.0 days for the original strain.¹⁴ Studies from Japan and South Korea reported similar findings. Japanese data placed the Omicron BA.5 incubation period at about 2.6 days, well below the Delta-era average.¹⁵ Korean researchers estimated a mean of 3.5 days for Omicron BA.1 versus 6.5 days during Delta dominance, and noted that 95 percent of symptomatic Omicron cases developed symptoms within six days of exposure.¹⁶

A shorter incubation period means the contagious window likely begins sooner after exposure. If you are exposed to a circulating Omicron subvariant, you could realistically be shedding virus within two to three days, rather than the four to five days that was typical earlier in the pandemic. This compressed timeline makes it harder to catch an infection before it spreads, but it also means the overall course from exposure to recovery tends to be shorter for most people.

People Who Stay Contagious Much Longer Than Average

While most people clear infectious virus within about a week of symptom onset, immunocompromised individuals are a major exception. A systematic review concluded that people with weakened immune systems can harbor replication-competent virus in their respiratory tracts for months, sometimes even while feeling well.¹⁷ A study focused on Omicron-era infections in immunocompromised patients found that the most prolonged culture-positive samples, those beyond 20 days, came from patients on B cell depletion therapy, a treatment that cripples the antibody response the body uses to clear the virus.¹⁸ Reassuringly, a separate study found that 90 percent of specimens collected more than 21 days after a positive PCR in immunocompromised patients did not contain viable virus. But the remaining ten percent did, and those cases are the ones that make blanket day-count rules unreliable for this group.¹⁹

Children present a different pattern. A systematic review of pediatric viral shedding found that children shed detectable RNA through the respiratory tract for an average of about 11 days from symptom onset, with the range extending up to 24 days. More unusually, 86 percent of children tested via stool or rectal swabs returned positive results, with gastrointestinal shedding lasting an average of roughly 24 days. In 89 percent of cases, GI shedding persisted after nose and throat swabs had already turned negative.²⁰ This extended gastrointestinal shedding is worth knowing about in settings like daycare, though fecal-oral transmission is not considered a major route compared to respiratory spread.

Antiviral Treatment and Viral Rebound

Antiviral drugs like nirmatrelvir-ritonavir (Paxlovid) can reduce the severity and duration of COVID symptoms, but they have also become associated with a frustrating phenomenon: viral and symptom rebound after treatment ends. A randomized clinical trial comparing two antivirals found that roughly 20 to 22 percent of participants experienced a rebound in viral RNA levels after completing treatment, and about a quarter had a recurrence of symptoms.²¹ The rebound rates were similar between the two drugs studied, suggesting the phenomenon is not unique to any one antiviral but rather reflects what happens when treatment suppresses the virus before the immune system has fully developed its own response.

If you experience a rebound, the practical question is whether you are contagious again. The evidence on this is still evolving, but the CDC advises treating a rebound as a potential return of infectiousness. If your symptoms come back or you test positive again on a rapid test after initially improving, you should restart the clock on precautions, including staying home while symptomatic and masking around others for the following days.

Using Rapid Tests to Judge Whether You Are Still Contagious

Given the disconnect between PCR results and actual infectiousness, rapid antigen tests have emerged as the most practical tool for deciding when you can safely be around others. A longitudinal cohort study of non-hospitalized adults with COVID found that the presence of N antigen (the protein detected by most rapid tests) was strongly associated with the ability to culture live virus. Between days six and ten after symptom onset, a positive rapid test corresponded to roughly a seven-to-eight-fold increase in the likelihood that infectious virus was still present, while neither viral RNA levels nor the presence of ongoing symptoms reliably predicted culture positivity during that window.²²

In other words, symptoms alone are an imperfect guide. You can feel mostly recovered and still be shedding live virus, or you can feel lousy from lingering inflammation while no longer being contagious. A rapid antigen test, while not perfect, threads the needle better than either a PCR (which stays positive too long) or symptoms (which are unreliable on their own). If you are trying to decide whether it is safe to visit a vulnerable relative or return to a shared workspace, a negative rapid test is the most actionable signal available outside of a research lab.

Why Isolation Compliance Matters and What Gets in the Way

All of this biological nuance is only useful if people actually stay home when they are contagious, and the evidence on compliance tells a more complicated story than “follow the guidelines.” Research using cellphone mobility data found that residents of lower-income areas in the US complied with shelter-in-place orders significantly less than those in wealthier areas, even after accounting for factors like political affiliation and population density. The 2020 CARES Act stimulus payments measurably increased social distancing in the communities that received them, suggesting the problem was economic as much as behavioral.²³

An Israeli survey made the connection even more starkly: when people were assured of financial compensation during self-quarantine, compliance reached 94 percent. Without that assurance, it dropped below 57 percent.²⁴ Individual personality traits also play a role. A study on economic preferences during the crisis found that people with higher risk tolerance were less likely to avoid crowds, while those with stronger senses of social responsibility were more likely to follow public health rules.²⁵ The takeaway here is that even a perfectly calibrated contagious-period guideline is only as effective as the conditions that allow people to follow it. Paid sick leave, workplace flexibility, and cultural norms around illness all determine whether the science of viral shedding translates into fewer infections in practice.

Children and Gastrointestinal Shedding

The extended gastrointestinal shedding in children, noted earlier, raises questions that go beyond respiratory spread. The finding that children can shed viral RNA in their stool for an average of about 24 days, often weeks after respiratory swabs turn negative, has prompted ongoing investigation into whether this represents a meaningful transmission route.²⁰ Fecal-oral spread has not been identified as a significant driver of COVID transmission in the general population, and most public health guidance does not factor it into isolation recommendations. But in environments with young children who may not have reliable hygiene practices, like childcare centers, the persistence of virus in stool is relevant context. It reinforces why hand hygiene and diaper-changing protocols are not just general good practice but have specific relevance during and after a COVID infection in a household with young kids.

What remains unclear is how often the RNA detected in stool represents live, replication-capable virus versus the same kind of dead fragments that confound respiratory PCR results. The few studies that attempted viral culture from pediatric stool samples had mixed results, and the area is understudied compared to respiratory shedding in adults. For practical purposes, the respiratory route remains the main concern, and a child who tests negative on a rapid nasal swab is unlikely to be a significant source of airborne spread, regardless of what might still be happening in their gut.