The honest answer is that, more than five years after the pandemic began, no one has definitively proven how SARS-CoV-2 first infected humans. What is known is that a cluster of unusual pneumonia cases appeared in Wuhan, China, in late 2019, and that the two leading explanations, a natural spillover from animals and a laboratory-associated incident, both remain plausible according to every major investigation conducted so far.
The First Known Cases in Wuhan
On December 31, 2019, the Wuhan Municipal Health Commission reported a cluster of pneumonia cases with no clear cause. Many of the earliest patients had links to the Huanan Seafood Wholesale Market, a large indoor market in central Wuhan that sold live animals alongside seafood and meat. A novel coronavirus was eventually identified as the pathogen, later named SARS-CoV-2.
Genetic sequencing showed the virus was a close relative of coronaviruses found in horseshoe bats in southern China. The closest known bat virus, called RaTG13, shares about 96.2% of its genome with SARS-CoV-2. That sounds high, but the remaining gap represents decades of evolutionary distance, meaning the pandemic virus did not jump directly from bats to people. Something happened in between, either passage through an intermediate animal host or some other chain of events that allowed the virus to adapt to human cells.
The Natural Spillover Hypothesis
Coronaviruses have a track record of jumping from animals to humans. The original SARS outbreak in 2003 was traced to civets sold in Chinese live-animal markets, and MERS coronavirus reached people through dromedary camels. The pattern is familiar: a bat virus circulates in a wildlife species that is handled, farmed, or sold in close contact with people, and at some point the virus gains the ability to infect human cells.
Environmental samples collected from the Huanan market in January 2020 support the idea that wildlife played a role. Out of 159 market samples that were later sequenced, six tested positive for SARS-CoV-2. Five of those six came from the same stall or its immediate vicinity, a stall identified as having sold live wildlife. Genetic analysis of those samples revealed abundant raccoon dog DNA mixed in with the viral material, particularly in and around that stall. Raccoon dogs are known to be susceptible to SARS-related coronaviruses.
The catch is that these samples were collected weeks after the outbreak was recognized, and by that point the market had been cleaned and shut down. No individual animal from the market has ever tested positive for the virus. So the evidence is circumstantial: the virus and animal DNA were found in the same place, but there is no smoking-gun sample proving an animal was infected first.
The Laboratory Incident Hypothesis
Wuhan is also home to the Wuhan Institute of Virology (WIV), one of China’s top research labs for bat coronaviruses. The lab had been collecting and studying these viruses for years, including work on RaTG13, which was sampled from a cave in Yunnan Province in 2013 after several miners fell ill with a SARS-like disease. The WIV also has a published record of conducting gain-of-function research, experiments that modify viruses to study how they might become more transmissible or more dangerous.
A U.S. State Department fact sheet released in early 2021 stated that several WIV researchers became sick in autumn 2019, before the first publicly identified COVID case, with symptoms that could be consistent with either COVID-19 or ordinary seasonal illness. The document also raised concerns that the WIV had not been transparent about its closest viral samples and had conducted classified research on behalf of the Chinese military since at least 2017.
Supporters of this hypothesis point out that a lab studying the very family of viruses responsible for the pandemic was located in the same city where the outbreak began. Critics counter that Wuhan is a city of 11 million people and a major transportation hub, and that the lab was studying bat coronaviruses precisely because they were known to pose a spillover risk in the region.
The Furin Cleavage Site Debate
One feature of SARS-CoV-2 became a flashpoint in the origins debate. The virus contains a small insertion in its spike protein, a structure called a furin cleavage site, that helps it enter human cells efficiently. No other known close relative of the virus has this feature, which makes it unusual among its family.
Some scientists argued this insertion looked engineered. The virologist David Baltimore initially called it a “powerful challenge to the idea of a natural origin,” though he later clarified that you cannot determine the virus’s origin just by looking at the genetic sequence. Others noted that similar cleavage sites have evolved naturally in more distantly related coronaviruses, so the feature is not biologically unprecedented.
What complicated the picture further was a 2018 research proposal, involving an international group of scientists that included WIV researchers, which outlined plans to identify novel cleavage sites in bat coronaviruses and insert them into related viruses to study their function. The proposal described a workflow that could produce a virus “practically indistinguishable from a rare” naturally occurring one. Whether that work was ever carried out remains unclear, but its existence made it harder for scientists to rule out either explanation based on genomics alone. As one researcher put it: “There is no way to know whether humans or nature inserted the site.”
What Investigations Have Concluded
The U.S. Intelligence Community released an unclassified assessment that laid out where its agencies stood. Four agencies and the National Intelligence Council assessed, with low confidence, that the pandemic most likely began through natural exposure to an infected animal. One agency assessed, with moderate confidence, that a laboratory incident was the more likely cause. Three agencies could not reach a conclusion either way. All agencies agreed the virus was not developed as a biological weapon, and most assessed that it probably was not genetically engineered.
The World Health Organization established a dedicated panel, the Scientific Advisory Group for the Origins of Novel Pathogens, to continue evaluating the evidence. As of mid-2025, that group has published frameworks for investigating pathogen origins and released an independent assessment, but China’s limited data sharing has remained a persistent obstacle. Without access to the WIV’s full database of viral sequences (which was taken offline in September 2019), raw patient records from early cases, or blood samples from market workers and lab staff, the question may never be fully resolved.
Why It Still Matters
The origins question is not purely academic. If the virus jumped from animals at a market, it strengthens the case for tighter regulation of wildlife trade and better surveillance at the human-animal interface. If a lab accident was the cause, it underscores the need for stricter international biosafety standards and independent oversight of high-risk research. Both risks are real regardless of which one sparked this particular pandemic, but knowing the answer would help governments and scientists prioritize where to invest in prevention.
What the evidence currently supports is a picture with two plausible paths and not enough data to choose definitively between them. The earliest known cluster centered on a market that sold live wildlife susceptible to the virus. The city also housed a laboratory doing exactly the kind of work that could, in theory, produce such a virus. Both facts are true, and neither has been sufficient to close the case.