Most people with long COVID recover within a year, but the timeline varies widely. Large tracking studies show that about 77% of infected individuals feel fully recovered by six months, rising to roughly 83% by one year. After two years, around 17% still report lingering symptoms. A small percentage, estimated at 5% or less, may experience symptoms that persist beyond three years.
Recovery Rates at 6, 12, and 24 Months
A major study published in The BMJ followed people after SARS-CoV-2 infection and found that 22.9% had not fully recovered at six months. That number dropped to 18.5% at 12 months and 17.2% at 24 months. The steepest improvement happens in the first six to nine months, with recovery slowing considerably after that. If you’re still symptomatic at the one-year mark, the odds of full resolution in the following year are lower but not zero.
A separate prospective study found that about 21.5% of people met criteria for “persistent long COVID” at the two-year mark, meaning they had at least one symptom that never fully went away. In a best-case analysis (assuming all people who didn’t respond to the survey were symptom-free), the floor was closer to 6%. Recent U.S. data suggests that 90% to 100% of people recover within three years, with roughly 5% still experiencing symptoms that don’t completely resolve.
How Specific Symptoms Resolve
Not all long COVID symptoms follow the same clock. Brain fog, one of the most distressing symptoms, typically clears between six and nine months after infection, according to Yale Medicine. In some people it lingers up to 18 months or longer, but complete resolution is the norm rather than the exception.
Physical symptoms like fatigue and reduced activity tend to be more stubborn. One study tracking recovery from hospital discharge found that physical symptoms (fatigue, shortness of breath, headache, sleep problems) were still present in 82% of patients after one year and 45% after 18 months, only dropping to about 10% after 20 months. Neuropsychological symptoms like anxiety, depression, and memory problems persisted in 59% of patients at one year and 28% at 18 months. The pattern is clear: physical fatigue often outlasts cognitive symptoms, but both continue improving well past the one-year mark.
What Affects How Fast You Recover
Your recovery speed depends on a combination of factors. A large German cohort study identified several characteristics tied to longer recovery times. People between ages 49 and 59 had a 30% lower chance of becoming symptom-free compared to those under 49. Women recovered more slowly than men. People with lower psychological resilience took significantly longer to clear their symptoms.
How sick you were initially matters a great deal. People who had six or more symptoms during their acute infection took about three times as long to reach the symptom-free milestone (a median of 21 days to first quartile recovery versus 7 days for those with fewer symptoms). By nine months, only 41.6% of the high-symptom group was symptom-free, compared to 52.3% of those who had a milder acute illness. Hospitalization during the initial infection, higher BMI, and pre-existing conditions like lung disease, cardiovascular disease, or neurological conditions also predicted a longer road to recovery.
Vaccination and Recovery
Getting vaccinated after infection appears to improve your chances of recovering from long COVID. A systematic review in BMJ Medicine found that studies on post-infection vaccination showed odds ratios for long COVID ranging from 0.38 to 0.91, meaning vaccinated individuals were less likely to still have symptoms. One study specifically looking at recovery from existing long COVID found that vaccinated patients had roughly half the odds of remaining unrecovered compared to unvaccinated patients. None of the studies found that vaccination made long COVID worse.
Reinfection Can Reset the Clock
Getting COVID again while recovering from long COVID is a real concern. A large study of over 465,000 children and adolescents in the RECOVER database found that reinfection roughly doubled the risk of a long COVID diagnosis (relative risk of 2.08). The increased risk wasn’t limited to one type of symptom. It spanned fatigue, cognitive impairment, heart-related issues, headaches, and mental health problems. While this data comes from pediatric populations, the implication is straightforward: avoiding reinfection during recovery matters.
Long COVID in Children
Children follow a somewhat different recovery curve. A one-year prospective study found that 85.8% of children with long COVID still had symptoms at three months, but recovery accelerated from there. By six months, 14% had fully recovered. By nine months, 43.9% were symptom-free. At one year, 67.5% no longer reported any long COVID symptoms. The most persistent symptoms in kids were fatigue (still present in 19.8% at one year), reduced physical activity (13.9%), and headache (12.3%). Children over age six were more likely to report memory problems, difficulty concentrating, and emotional instability alongside their physical symptoms.
What “About Half Improve” Looks Like in Practice
A University of Iowa long COVID clinic enrolled 403 adults and tracked their progress. About 51% reported meaningful subjective improvement since their long COVID began. That’s an important number to sit with: it means roughly half of people who sought care at a specialized clinic felt significantly better, but also that about half did not. Improvement and full resolution are different things. Many people with long COVID experience a gradual reduction in symptom severity and frequency rather than a single moment where everything clears. You might find that your worst days become less frequent, your energy ceiling slowly rises, or symptoms that were constant become intermittent.
The overall picture is one of slow, uneven progress. The first year brings the most recovery. The second year brings smaller but real gains. Beyond two years, a small minority still has persistent symptoms, though even in this group, complete resolution remains possible.

