Bats almost certainly harbor the ancestor of SARS-CoV-2, the virus that causes COVID-19, but whether the virus jumped directly from bats to humans or passed through another animal first remains unresolved. The closest known relative of SARS-CoV-2 is a bat coronavirus called RaTG13, which shares 96% of its genetic sequence with the pandemic virus. That 4% gap represents decades of evolutionary divergence, meaning SARS-CoV-2 did not come straight from any bat virus scientists have yet identified. A June 2025 report from the WHO’s Scientific Advisory Group concluded that “the weight of available evidence suggests zoonotic spillover, either directly from bats or through an intermediate host,” while noting that all hypotheses, including a laboratory leak, remain on the table.
Why Bats Are the Prime Suspects
Bats are not new to this story. In the past two decades, three coronaviruses with ancestral origins in bats have spilled over into humans: the original SARS virus in 2002, MERS, and now SARS-CoV-2. At least five of the seven coronaviruses known to spread between people trace their ancestry back to bat coronaviruses, including two that now circulate as common cold viruses.
The specific bats in question are horseshoe bats, a group found across Asia, Europe, and Africa. The closest viral relatives to SARS-CoV-2 have been sampled from several horseshoe bat species: Rhinolophus affinis in China’s Yunnan province (where RaTG13 was collected in 2013) and multiple species in northern Laos, where viruses called BANAL-103 and BANAL-236 were found in 2020. These viruses are genetically close to SARS-CoV-2 but not close enough to be its direct parent.
How Bat Viruses Gain the Ability to Infect Humans
SARS-CoV-2 infects human cells by latching onto a protein called ACE2 on the cell surface. Research published in Nature has shown that the ability to bind ACE2 is actually an ancient trait among this family of coronaviruses, not something that had to be invented from scratch. Many bat coronaviruses already bind ACE2 in their bat hosts, and because ACE2 varies between bat species, these viruses are constantly adapting to new versions of the protein.
The key finding is that this adaptation can happen remarkably easily. In laboratory experiments, researchers found that single amino acid changes at just a handful of positions on the virus’s surface protein could enable binding to human ACE2. In the majority of cases where a virus already bound one version of ACE2, there were single mutations that could improve binding to a new version by more than fivefold. This means the evolutionary path from infecting bats to infecting humans is shorter than it might seem, and there are multiple routes a virus could take to get there.
The Missing Link: An Intermediate Host
When SARS emerged in 2002, the virus passed from horseshoe bats into farmed Himalayan civets before reaching humans. MERS took a similar path, moving from bats into dromedary camels likely decades before it began infecting people. This pattern of bat-to-intermediate-animal-to-human transmission is well established, and many researchers expected the same for SARS-CoV-2.
Several candidate species have been investigated. Pangolins were an early focus because coronaviruses related to SARS-CoV-2 had been found in them. Raccoon dogs drew attention because they were implicated in the 2002 SARS outbreak and are commonly traded in Chinese wildlife markets. Both species are susceptible to the virus in laboratory settings.
The problem is that no infected animal has been found. Environmental sampling at the Huanan Seafood Market in Wuhan, where many early cases clustered, detected SARS-CoV-2 in 74 environmental samples (surfaces, drains, equipment) but in none of the 457 animal samples tested across 18 species. The market did sell live wildlife from 10 stalls, including snakes, bamboo rats, hedgehogs, and other species, but researchers have not been able to confirm viral infection in any of them. Sale records show no bats were sold at the market.
Decades of Hidden Circulation
The 96% genetic match between SARS-CoV-2 and RaTG13 sounds high, but in viral evolution terms it represents a long separation. Molecular clock analyses estimate that the lineage leading to SARS-CoV-2 diverged from its closest known bat relatives somewhere between the 1940s and the early 2000s, with estimates centering around 1969. This means the virus’s ancestors were circulating undetected in bats, and possibly other animals, for roughly 40 to 70 years before the pandemic.
During that time, the virus would have continued evolving, potentially passing between different bat species or other wildlife. Studies of villagers in Yunnan province have found low levels of antibodies against SARS-related coronaviruses in people with no history of SARS infection, suggesting that small, unnoticed spillover events from bats to humans have been happening for years without triggering outbreaks.
What Remains Unanswered
The honest answer is that the exact chain of transmission from bats to humans has not been proven. No one has found the direct progenitor virus in any animal. No intermediate host has been confirmed. The environmental evidence from the Huanan market places the virus there early in the outbreak but cannot prove it originated there. And China has not provided full access to early samples, patient data, or laboratory records from the Wuhan Institute of Virology, which studied bat coronaviruses.
What the genetic and ecological evidence does show convincingly is that bats carry a vast reservoir of coronaviruses closely related to SARS-CoV-2, that these viruses are pre-adapted to bind human cells with minimal evolutionary change, and that bat-to-human spillover (with or without an intermediate host) has caused previous coronavirus outbreaks. The WHO’s advisory group considers this the most likely explanation for the pandemic’s origin, while acknowledging that certainty may never be reached without fuller cooperation and data sharing from the countries involved.