Is COVID Airborne? What the Science Shows About Transmission

SARS-CoV-2, the virus that causes COVID-19, spreads primarily through the air. Researchers have recovered viable virus from air samples collected several meters from infected patients, documented transmission patterns in real-world outbreaks that only airborne spread can explain, and shown in animal models that the virus travels between hosts through shared air even when direct contact is impossible. This conclusion was not universally accepted at the start of the pandemic, largely because of a decades-old error in how public health authorities categorized respiratory transmission. The full picture of how COVID travels through air, how long it remains dangerous, and what actually reduces the risk is more detailed than “it’s airborne” implies.

A Century-Old Mistake About Droplets

For most of the twentieth century, public health doctrine held that respiratory infections spread mainly through large droplets that fall to the ground within a meter or two of the infected person. This framework traces back to 1910, when a prominent public health official named Charles Chapin argued that close-range “sprayborne” droplets were the main route and that airborne transmission was very unlikely. His reasoning became the dominant paradigm, and for over fifty years airborne spread was dismissed as unimportant for nearly all respiratory diseases. That only began to change in 1962, when researchers proved that tuberculosis, long assumed to spread by droplets, was in fact airborne.¹

A key piece of the old framework was a size cutoff: particles smaller than 5 micrometers were called “aerosols” (capable of lingering in the air), while anything larger was called a “droplet” (assumed to fall fast). Both the WHO and the CDC used this 5 µm threshold in their early COVID-19 guidance. But the cutoff was never scientifically grounded for determining how far particles travel. It actually originated from research on which particle sizes penetrate deepest into the lungs, not on which ones stay suspended in the air. Particles well above 5 µm can float for minutes or longer in typical indoor conditions, and the conflation of “what reaches the lower lungs” with “what stays airborne” muddled the pandemic response for months.²

The practical consequence was that early pandemic guidance emphasized handwashing, surface disinfection, and maintaining a six-foot distance, while downplaying the importance of ventilation and high-quality masks. These measures were not wrong, but they reflected an incomplete picture of how the virus actually traveled.

What Superspreading Events Revealed

Some of the strongest evidence for airborne transmission came from outbreaks that simply could not be explained by close-range droplets or contaminated surfaces. A Washington State church choir rehearsal, a Korean call center, a Korean exercise class, and two Chinese bus trips were all studied in detail. In each case, the likely source patient was presymptomatic or only mildly symptomatic, and infections reached people seated far from the source. Modeling these events as airborne transmission produced consistent estimates of the number of virus copies needed to establish infection, roughly in the same range across all five cases.³

Choir rehearsals became a recurring theme. An analysis of two choir outbreaks in Berlin in 2020 found that about four out of five infections were attributable to long-range transmission, meaning the infected singers were not close enough to the index case for large droplets to be the explanation. The combination of singing for extended periods in enclosed spaces with poor ventilation created ideal conditions for aerosol buildup.⁴

A detailed investigation of two buses that carried passengers from the same departure point reinforced this picture. One bus had the presymptomatic index case aboard; the other did not. The bus with the infected passenger saw widespread transmission even among people seated several rows away, while fomite transmission played a negligible role. The researchers concluded that airborne spread dominated in this poorly ventilated indoor space.⁵

Live Virus Captured from Room Air

Epidemiological detective work can tell you that people got infected across a room, but it cannot prove the virus was floating in the air between them. For that, you need to catch it. Researchers in a hospital room with COVID-19 patients collected air samples at distances ranging from 2 to nearly 5 meters from the patients and isolated viable SARS-CoV-2. Genome sequencing confirmed the airborne virus matched the strain from a newly admitted patient. The estimated concentrations were modest but real, ranging from 6 to 74 infectious units per liter of air.⁶

This finding was significant because early skeptics argued that even if viral RNA could be detected in air, the virus might not be infectious. The hospital study showed it was. Meanwhile, animal experiments using hamsters provided controlled confirmation: SARS-CoV-2 passed between animals housed in separate cages connected only by shared airflow. When a surgical mask barrier was placed between the cages, transmission dropped.⁷

How Long the Virus Survives in Air

Once expelled, SARS-CoV-2 does not stay dangerous indefinitely. Its survival in aerosol form depends heavily on temperature, humidity, and sunlight. Under conditions mimicking a cool, dark indoor environment, the time needed for a 90% drop in infectious virus exceeded two hours. At the other extreme, under warm temperatures with strong simulated sunlight, that same 90% decay happened in under five minutes.⁸

Temperature and humidity also matter independently of sunlight. The virus survives best at low temperatures and extreme humidity levels. At around 10°C and 40% relative humidity, the estimated half-life exceeded 24 hours. At 27°C and 65% relative humidity, it was closer to an hour and a half. These findings help explain why outbreaks clustered in cool indoor settings like meat-packing plants, where low temperatures and recirculated air created near-ideal conditions for the virus to persist.⁹

The chemical composition of the respiratory droplet itself also plays a role. When expelled droplets contain high levels of proteins (as human respiratory fluid does), those proteins can form a glassy gel-like shell around the virus as the droplet dries, shielding it from inactivation, particularly under low humidity. This helps explain why virus survival in real respiratory aerosols does not always follow the predictions based on simple water-droplet experiments.¹⁰

Why Some People Generate More Aerosol Than Others

Not everyone produces the same amount of infectious aerosol. The activity matters enormously. Measured in particles per second, speaking produces far fewer aerosols than singing, and singing produces far fewer than shouting. In one study of adolescents, speaking generated roughly 16 to 267 particles per second, singing produced 141 to 1,240, and shouting launched 683 to over 4,300.¹¹ This is why choir rehearsals, exercise classes, and indoor worship services repeatedly showed up as superspreading contexts: these are settings where people vocalize loudly for extended periods in shared air.

Age also matters. Pre-adolescent children produce substantially fewer aerosol particles than adults. Across breathing, speaking, singing, and shouting, emission rates in 8-to-10-year-olds were roughly a quarter of adult levels, and the total volume of emitted particles was even lower.¹² This is one of the factors that may have contributed to generally lower observed transmission in primary school settings compared to adult workplaces, though it is far from the only variable at play.

Beyond activity and age, individual physiology makes a difference. A study of 146 people found that those with higher body mass index and older age exhaled significantly more aerosol particles, even during quiet breathing. People with the highest combined BMI-and-age scores exhaled meaningfully more particles than those with the lowest.¹³ Sex was not a significant factor. This kind of variation helps explain why, even in similar settings, some infected individuals seem to transmit the virus to many people while others infect nobody.

How Surface Transmission Stacks Up

Early in the pandemic, people wiped down groceries and let mail sit for days. The fear of surface (fomite) transmission was understandable but turned out to be disproportionate. The bus-trip study mentioned earlier quantified this directly: in a poorly ventilated enclosed space with a confirmed index case, airborne spread accounted for the vast majority of infections while fomite transmission was negligible.⁵ Surfaces can carry the virus, and hand hygiene still matters as a general infection-control measure, but the primary battleground for COVID transmission is the air you breathe, not the surfaces you touch.

The Six-Foot Rule and Its Limits

Social distancing guidelines built around staying six feet apart assumed that large droplets are the main threat and that they fall to the ground within that radius. For aerosols small enough to stay suspended and mix throughout a room, distance alone offers limited protection. An analysis from MIT found that the six-foot rule provides “little protection from pathogen-bearing aerosol droplets sufficiently small to be continuously mixed through an indoor space.” Direct respiratory jets from unmasked individuals can carry virus-laden particles well beyond six feet before they dilute into the room’s overall air.¹⁴

Distance still reduces risk in the short range, where the exhaled plume is most concentrated. But in an enclosed room with poor ventilation and extended exposure, the entire room becomes the risk zone, not just the area within two meters of the infected person. Time and ventilation matter at least as much as distance.

What Ventilation Can and Cannot Do

Opening windows is one of the simplest ways to reduce airborne virus concentrations indoors. Cross-ventilation, where air flows through the space from openings on opposite sides, is dramatically more effective than single-sided ventilation. In modeling of a 100-square-meter space, cross-ventilation could reduce a viral load from 10,000 copies to zero within 15 minutes, while single-sided ventilation cut it only in half at best.¹⁵

Mechanical ventilation, the kind most office buildings use, introduces a counterintuitive wrinkle. Mixing-type ventilation systems work by blending fresh outdoor air with indoor air to dilute contaminants. Over long periods this helps. But a computational fluid dynamics study found that in the short term, within the first few minutes after a cough, increasing the ventilation rate in a mixing-type system actually increased virus exposure at distances up to 3 meters. This happened because the turbulent mixing pushed the concentrated plume outward before it had time to dilute. The effect faded over time, which is why general guidance to increase ventilation remains valid for sustained exposure. But it does highlight that ventilation is not a magic button: how the air moves matters, not just how much of it moves.¹⁶

Organizations like ASHRAE and the CDC have since published guidelines for controlling infectious aerosols in buildings, recommending measures ranging from upgraded HVAC filtration to adding air-cleaning technologies.¹⁷

Masks, Air Cleaners, and Filtration

If the virus travels through air, then filtering that air at the source, in the room, or at the recipient’s nose and mouth all reduce transmission. For masks, the material and layering matter considerably. Single layers of common fabrics ranged wildly in filtration efficiency, from as low as 5% to as high as 95% depending on the fabric and particle size. But hybrid combinations of different fabrics, like cotton layered with silk or chiffon, achieved above 80% filtration for very small particles and above 90% for larger ones. The likely reason is that combining a tightly woven fabric with one that carries a static charge provides both mechanical and electrostatic filtering.¹⁸ Of course, fit matters just as much as material. A high-efficiency mask worn loosely around the cheeks lets air bypass the filter entirely.

Portable HEPA air cleaners and far-UVC light systems offer room-level filtration. Testing against a viral surrogate in a large chamber showed that both technologies performed meaningfully, and their effectiveness roughly doubled when the aerosolized virus was carried in simulated saliva rather than pure water, suggesting that lab tests using water alone may underestimate how well these devices perform in real settings with actual respiratory particles.¹⁹

Why Omicron Spread Faster Through the Air

Not all SARS-CoV-2 variants shed equally into the air. In hospitalized patients, air samples were positive for the Omicron variant in about 55% of cases, compared to only 15% for Delta. Both the variant itself and the patient’s nasopharyngeal viral load were independently associated with whether virus could be detected in the surrounding air. Prior vaccination did not significantly affect air sample positivity.²⁰ This difference in airborne shedding is one plausible explanation for why Omicron spread so much faster in the population than earlier variants, on top of its immune-evasion properties.

Hamster experiments with the Delta variant showed that airborne transmission could occur with as little as one hour of shared airflow, though 15 minutes was generally insufficient. Transmission efficiency in hamsters did not vary significantly between temperate and tropical environmental conditions, suggesting that host-side and viral factors may matter more than ambient temperature and humidity when exposure is brief and direct.²¹

Hospital Procedures and the “Aerosol-Generating” Label

During the pandemic, certain medical procedures were flagged as “aerosol-generating” and required extra precautions, including full airborne-level personal protective equipment. Intubation, the process of placing a breathing tube in a patient’s throat, was near the top of that list. But measurement studies found that the aerosol produced during standard tracheal intubation was extremely low, far less than a single cough. Extubation (removing the tube) produced more aerosol, especially if the patient coughed during the process, but even then it fell well below what a voluntary cough generates.²²

A separate study found that mask ventilation during general anesthesia generated the highest peak aerosol concentrations among the procedures tested, and that uncomplicated intubations produced aerosol concentrations comparable to coughing. Interestingly, difficult intubations, which one might expect to generate more aerosol because of more manipulation, actually produced significantly less aerosol than uncomplicated intubations.²³ These findings have prompted calls to rethink which procedures truly warrant the aerosol-generating label and which were swept into the category more by tradition than by data.

Where Particles End Up in the Lungs

The size of an inhaled particle determines where it lands in your respiratory tract, and that has real consequences for disease severity. Large particles (above roughly 6 µm) tend to deposit in the upper airways, the nose and throat. Very small particles, below about 2 µm, bypass the upper tracts and settle deep in the alveoli, the tiny air sacs where oxygen exchange occurs.²⁴ Particles in the mid-range penetrate to the central and small airways. The smallest aerosols, the kind that stay airborne longest and travel farthest, are precisely the ones that can reach the deepest lung tissue. This is why aerosol transmission of COVID-19 is not just an academic distinction: the infection it produces when inhaled as fine aerosol may establish itself deeper in the lungs than when caught from a large droplet at close range.

Ultra-fine particles below 100 nanometers behave somewhat differently, tending to deposit in the upper airways again due to their high diffusion rates.²⁵ The upshot is that there is a “sweet spot” of particle sizes, roughly 0.1 to 5 µm, that are both small enough to remain airborne for extended periods and in the right range to deposit efficiently in the lower lungs. SARS-CoV-2-laden aerosols often fall squarely in this window.