Long COVID in 2026: Latest Research, Treatments, and What Scientists Have Learned

More than five years after the pandemic began, Long COVID has settled into an uncomfortable place in modern medicine: well-documented, still poorly understood, and — until very recently — largely untreatable by anything more than symptom management. But 2026 has brought a shift. After years spent simply describing the condition, researchers are now publishing hard data on whether specific treatments actually work. The results are a mix of disappointment, cautious optimism, and a lot of open questions.

The scale of the problem

Long COVID — also called post-acute sequelae of SARS-CoV-2 infection, or PASC — is generally defined as symptoms that persist for at least three months after infection. Estimates of how common it is vary widely depending on the population studied, but current figures suggest it affects a substantial share of people who've had COVID-19, with millions of Americans still living with persistent symptoms such as fatigue, brain fog, exercise intolerance, and autonomic dysfunction. A 2026 NASEM consensus effort has worked to sharpen the clinical definition, which has historically been a barrier to consistent diagnosis and research.

What's been learned

The biggest research effort here is the NIH's RECOVER initiative, a roughly $1.15 billion program that has enrolled tens of thousands of adults and children. Its early years focused on characterizing the condition; 2026 marked a turn toward testing actual treatments.

A few key findings have emerged:

• Long COVID doesn't follow one path. Research has identified several distinct symptom trajectories — some people have consistent symptoms from the start, while others don't develop problems until months after their initial infection.
• Vaccination appears protective, including in teens. An observational study found adolescents who'd been vaccinated in the six months before their first COVID-19 infection were roughly a third less likely to develop Long COVID, adding to earlier evidence of a similar protective effect in adults.
• Two once-promising drug repurposing candidates didn't pan out. A RECOVER trial testing extended courses (15 and 25 days) of the antiviral Paxlovid — based on the theory that a lingering viral reservoir might be driving symptoms — found no meaningful improvement in fatigue-related symptoms, cognitive dysfunction, or orthostatic intolerance compared with a short course. Metformin, which had shown some promise for preventing Long COVID when given during acute infection, also failed to help once Long COVID was already established.
• Cognitive rehab approaches showed only modest, non-specific benefit. The RECOVER-NEURO trial — the first major clinical trial result from the program, published in JAMA Neurology — tested computerized brain-training software, structured cognitive rehabilitation, and transcranial direct current stimulation against comparison groups in 328 participants with brain fog and related cognitive symptoms. All five study arms showed mild improvement over time, but no single intervention outperformed the others, meaning none could be shown to work better than a comparison condition.

Taken together, this is real progress in the sense that it's ruling things out — but as of mid-2026, no drug has FDA approval specifically for Long COVID.

What shows promise

The picture isn't purely negative. Several avenues are still active and, in some cases, encouraging:

• RECOVER-TLC (Treating Long COVID), a newer phase of NIH-backed trials run with the Foundation for the National Institutes of Health, is currently testing baricitinib (an anti-inflammatory arthritis drug), low-dose naltrexone, semaglutide (a GLP-1 drug), and stellate ganglion block — a nerve-blocking procedure sometimes used for chronic pain and autonomic symptoms. Results are expected to roll out through the rest of 2026.
• Biomarker research is advancing. Small studies have found signs of ongoing inflammation and altered markers of brain plasticity in people with persistent cognitive symptoms compared with those who've fully recovered, which could eventually support an objective test for Long COVID rather than relying purely on self-reported symptoms.
• Autonomic dysfunction is getting more attention as a treatable target. A notable share of severely affected patients meet diagnostic criteria for POTS (postural orthostatic tachycardia syndrome), a condition affecting blood flow regulation that has established, if imperfect, treatment approaches of its own — offering a more concrete path forward for at least a subset of patients.
• Some claims of "breakthrough" therapies should be treated with real skepticism. Marketing-oriented health sites have promoted specific monoclonal antibody products for Long COVID as newly approved breakthroughs; these claims aren't reflected in the peer-reviewed literature or NIH trial data, and readers should be wary of treatments advertised this way without backing from bodies like the FDA or a published clinical trial.

What's still unknown

• No universally accepted biological mechanism. Leading theories include persistent viral reservoirs, autoimmune-like inflammation, microclotting, and nervous-system dysregulation — but none has been definitively proven as the primary driver, and different mechanisms may explain different patients' symptoms.
• No validated diagnostic biomarker exists yet. Diagnosis still relies mainly on symptom history and ruling out other causes.
• Why some people develop Long COVID and others don't remains unclear, despite known risk factors like severity of initial infection, certain pre-existing conditions, and vaccination status.
• Long-term prognosis is murky. Some patients improve significantly over one to two years; others report ongoing quality-of-life impairment more than two years after infection, and researchers don't yet have a reliable way to predict, at the time of infection, who will fall into which group.

The bottom line

Five-plus years in, Long COVID research has moved from description to experimentation — and while several early hopes (extended Paxlovid, metformin, and most cognitive-rehab approaches) haven't held up under rigorous testing, a genuine pipeline of trials is now underway that should produce clearer answers on drug treatments within the next year or two. For patients, that means the most reliable path today is often symptom-targeted management — particularly for autonomic symptoms like POTS — in partnership with a knowledgeable clinician, while the next wave of RECOVER-TLC results plays out.