COVID-19 spreads primarily through virus-laden particles expelled from an infected person’s nose and mouth during breathing, talking, coughing, and sneezing. These particles range from large droplets that fall quickly to tiny aerosols that can linger in the air for minutes to hours, and the aerosol route has turned out to be far more important than early pandemic guidance suggested. The story of how the virus gets from one person to the next involves a surprising number of factors, from how loudly someone speaks to how well a room is ventilated, and some of the early assumptions about transmission turned out to be wrong.
Aerosols and Droplets
Every time you exhale, talk, cough, or sneeze, you release a turbulent cloud of gas that carries clusters of tiny liquid particles outward from your mouth and nose. These particles exist on a spectrum of sizes. Larger droplets, roughly above 50 to 100 micrometers, behave like tiny projectiles: gravity pulls them down quickly and they tend to land on nearby surfaces within a meter or two. Smaller aerosol particles, especially those under about 10 micrometers, behave more like smoke. They stay suspended in air, drift with currents, and can spread throughout an entire room. Modeling work has confirmed that particles around 1 and 10 micrometers remain airborne and disperse throughout an indoor space, while larger ones deposit on the floor relatively fast.
Early in the pandemic, public health messaging focused heavily on large droplets and recommended staying about two meters apart. That advice was not wrong, exactly, but it was incomplete. Research increasingly showed that the fine aerosols people produce during normal breathing and conversation carry enough virus to infect others, even across a room with poor ventilation. The scientific community eventually reached consensus that airborne aerosol transmission is the dominant route for COVID-19 spread indoors.
Why Talking, Shouting, and Singing Matter So Much
One of the more striking findings from transmission research is how dramatically vocal activity affects the number of particles a person puts into the air. Speaking loudly produces roughly ten times more particles than speaking quietly, with the size of those particles staying about the same regardless of volume. That means the total volume of respiratory fluid launched into the air goes up substantially when someone raises their voice.
Singing takes this even further. Measurements of professional classical singers found emission rates ranging from about 750 to over 6,000 particles per second, with a median around 1,500. Compared to speaking, singing produced a median enhancement of about 17 times more particles, with some singers hitting nearly 100 times more. The effect was driven mainly by louder sound pressure, not by the type of voice. This helps explain why choir rehearsals, karaoke bars, and nightclubs became notorious settings for superspreading events. It is not that any one activity is uniquely dangerous; it is that louder and more sustained vocalization pumps dramatically more potentially virus-carrying aerosol into the shared air.
When an Infected Person Is Most Contagious
The timing of viral shedding plays a critical role in how easily COVID-19 spreads. Viral load in the nose and throat tends to peak in the early symptomatic period, generally around four to five days after symptoms start, though some studies found peak loads right at the time of symptom onset or even slightly before. Infectious virus can be isolated from respiratory specimens during roughly the first eight to ten days of symptoms, after which the probability of recovering live virus drops quickly.
What made COVID-19 so difficult to control is that people are infectious before they feel sick. Research estimated that about half of secondary transmissions happen during the presymptomatic phase, when an infected person has no idea they are carrying the virus. This pre-symptomatic spreading is a major reason why temperature checks and symptom-based screening never worked well as standalone control measures. The presence of symptoms is simply an unreliable indicator of whether someone can pass the virus to others.
Asymptomatic Versus Presymptomatic Spread
People who never develop symptoms at all (truly asymptomatic cases) do spread the virus, but at a meaningfully lower rate than those who are presymptomatic or symptomatic. A systematic review and meta-analysis found that the transmission rate from asymptomatic index cases was about 1.8 per 100 person-days, compared to roughly 5 per 100 person-days for both presymptomatic and symptomatic cases. Another meta-analysis estimated asymptomatic individuals had about one-third the transmission risk of symptomatic ones. The distinction matters: presymptomatic people, who will eventually develop symptoms but have not yet, appear to spread the virus at rates similar to symptomatic people. Truly asymptomatic carriers are less efficient spreaders, but they are harder to identify, which partially compensates for their lower per-contact transmission rate.
Superspreading and the 80/20 Pattern
COVID-19 does not spread evenly from person to person. The transmission pattern is heavily “overdispersed,” meaning a small fraction of infected people cause a disproportionately large share of all new infections. Estimates suggest that roughly 10 to 15 percent of cases are the source of about 80 percent of infections, while the majority of infected people pass the virus to few or no others.
Analysis of contact-tracing data from Hong Kong and India illustrated this pattern clearly: depending on the modeling approach, about 10 to 31 percent of cases accounted for the vast majority of all onward transmission. This overdispersion gives the virus what researchers described as an “Achilles’ heel.” Reducing contacts between people who do not regularly meet, the kind of random mixing that happens at large gatherings, conferences, or crowded bars, has a much bigger effect on slowing the pandemic than reducing contacts within defined social groups like households or small workplaces. The virus needs those rare, high-transmission events to sustain itself at a population level.
Where Transmission Happens Most
Households are one of the best-studied transmission settings. Pooled estimates of the household secondary attack rate, the probability that an infected household member passes the virus to another member, land in the range of about 17 to 18 percent. That means if someone in your home gets COVID-19, each other household member has roughly a one-in-five or one-in-six chance of getting infected. The rate tends to be higher among older adult contacts and among contacts of symptomatic index cases. Spouses of index cases face about two to three times the risk of other household members, likely because of closer and more prolonged shared exposure.
Outside the home, the settings that produce the most transmission share common features: enclosed spaces, poor ventilation, lots of people, extended time together, and activities that increase aerosol production like talking loudly, singing, or heavy breathing during exercise. Restaurants, bars, religious services, gyms, and workplaces with open-plan seating and inadequate air exchange have all been linked to clusters. Outdoor settings, by contrast, feature essentially unlimited ventilation, which disperses aerosol rapidly and makes transmission much less likely.
How Indoor Air Quality Changes Risk
Ventilation is one of the most important factors determining whether an indoor environment is low-risk or high-risk for COVID-19 transmission. Increasing the rate at which fresh outdoor air replaces indoor air dilutes any virus-carrying aerosol in the room. A chamber study with infected participants demonstrated that increasing ventilation or running a portable HEPA air cleaner resulted in measurably lower concentrations of virus in the air. The relationship between risk and time indoors is not just about distance from an infected person; it depends on the cumulative exposure time, which itself depends on the room’s ventilation rate, its size, the number of occupants, their breathing rate, their activity level, and whether anyone is wearing a mask.
Temperature and humidity also play a role, though their effects have been somewhat less dramatic than many initially hoped. Animal studies found that high temperature combined with high humidity moderately slowed airborne transmission, partly by shortening the virus’s survival time in the air and partly by reducing viral replication. But the effect was less pronounced than the expectation that summer heat would substantially block spread. One fluid dynamics study found that virus viability dropped most when both high temperature and low relative humidity occurred together, but if humidity stayed high, the droplet cloud’s travel distance and concentration remained significant regardless of temperature. The practical takeaway: warm weather helps a little, but it does not make indoor transmission safe on its own.
Surface Transmission Turned Out to Be Minor
In the early months of the pandemic, intense focus was placed on surface contamination, or “fomite” transmission. People wiped down groceries, quarantined mail, and disinfected every doorknob. While SARS-CoV-2 RNA can persist on surfaces for days, detecting viral genetic material on a surface does not mean the virus is still infectious. Studies demonstrated that the RNA of common respiratory viruses can linger on surfaces long after the virus has lost its ability to infect cells, so detecting RNA alone cannot tell you much about actual transmission risk.
Research in real homes of infected people found that SARS-CoV-2 RNA contamination on surfaces was substantially reduced within about three days, and even that residual RNA likely had low or no infectivity. The consensus that emerged is that fomite transmission plays a limited role in COVID-19 spread. Handwashing remains good general hygiene, but the obsessive surface disinfection of 2020 was addressing a minor transmission pathway while the more important airborne route was underappreciated.
Can the Virus Enter Through the Eyes?
The receptor that SARS-CoV-2 uses to enter cells, called ACE2, is present on the surface of the eyes, including in the conjunctiva and cornea. Laboratory studies confirmed that these ocular surface cells express both ACE2 and a helper protein the virus needs, making them theoretically susceptible to infection. Some healthcare workers who treated COVID-19 patients without eye protection did develop infections, raising the possibility that virus-containing aerosol or droplets landing on the eyes could serve as an entry point.
However, ACE2 expression on the eye surface is much lower than in lung and kidney tissues, and the binding capability of the eye’s ACE2 to the virus’s spike protein is considerably weaker than in lung tissue. While the ocular route cannot be completely ruled out, especially in high-exposure settings like hospitals, it does not appear to be a meaningful driver of community transmission. The nose and upper airway remain the primary entry points. ACE2 expression is actually highest in the nasal lining, particularly in ciliated cells, and nasal cells are the primary targets for early SARS-CoV-2 replication.
An Unusual Route Through Plumbing
SARS-CoV-2 RNA has been detected in the feces of infected people, and researchers have investigated whether the fecal-oral route could contribute to spread. One particularly well-documented case involved infections among families living in vertically aligned apartments in a high-rise building in China. The infections and positive environmental samples were consistent with virus-laden aerosols traveling upward through shared drainage pipes in the bathrooms. This was not traditional fecal-oral transmission from contaminated food or water, but rather a hybrid route where aerosolized particles from sewage infrastructure entered living spaces through dry drain traps.
Events like this appear to be rare and confined to buildings with specific plumbing defects. They do not change the overall picture that respiratory aerosol during face-to-face or shared-room contact is how the vast majority of COVID-19 transmission occurs. But they underscore that building infrastructure can create unexpected pathways for airborne pathogens.
How Masks Reduce What Gets Into the Air
Masks work in two directions: they filter some of what the wearer breathes in (protection) and, often more effectively, they reduce what an infected wearer sends out (source control). In a controlled study where people actually infected with SARS-CoV-2 breathed through different face coverings, a properly worn duckbill N95 respirator reduced the exhaled viral load by about 98 percent, even without fit testing or training. Cloth masks and surgical masks also significantly reduced exhaled virus, with cloth masks in that study actually outperforming the tested surgical mask and a KN95.
The range of performance across mask types is wide. Tests of 15 different cloth mask designs found source-control collection efficiencies ranging from about 17 to 71 percent for cough-sized aerosols. Fit matters as much as the fabric itself: air that leaks around the edges of a mask bypasses the filter material entirely. N95 respirators, when properly sealed, offer the highest level of both source control and wearer protection, but even a simple cloth mask meaningfully reduces the amount of virus an infected person puts into a shared room.
Why Some People Produce More Aerosol Than Others
Not everyone generates the same amount of respiratory aerosol, and the variation is enormous. An observational study of nearly 200 healthy subjects found that exhaled particle counts varied between individuals by three orders of magnitude, a factor of roughly a thousand. Both COVID-19 infection itself and higher body mass index were associated with increased exhaled particle numbers.
Age is another significant factor. Older adults emit substantially more aerosol particles than younger people during both rest and exercise. One study found that subjects aged 60 to 76 emitted more than twice as many particles per minute as those aged 20 to 39, and about five times as much particle volume. A separate analysis found that age was the single most important predictor of small exhaled particle concentration, with the count roughly doubling over a 30-year period in adults. Sex and body mass index had minimal effects in that study. This age-related increase in aerosol production may be one factor, among many, contributing to the observation that older adults tend to be more efficient transmitters and are also more susceptible to infection.
How Viral Variants Changed the Transmission Equation
The version of SARS-CoV-2 that circulated in early 2020 was not the same virus that drove later waves. Each successive variant of concern evolved changes in its spike protein that affected how tightly and efficiently it bound to human cells. The Omicron variant, which emerged in late 2021, carried an unusually high number of mutations. Structural analysis revealed that Omicron’s spike protein formed new molecular bonds with the ACE2 receptor that compensated for other mutations known to weaken binding, resulting in binding strength similar to the Delta variant. Combined with increased ability to evade antibodies from prior infection and vaccination, these properties contributed to Omicron’s explosive spread.
Sub-variants continued to refine this process. Structural comparisons of Omicron sub-lineages like BA.1, BA.1.1, and BA.2 showed that even single mutations far from the receptor-binding region could enhance binding affinity through long-range changes in protein shape, a phenomenon researchers compared to a “butterfly effect” within the protein structure. The practical result was that each new wave tended to be driven by a variant that spread more easily than its predecessor, making the same indoor environments and exposure scenarios progressively riskier over time.
When the Virus Jumps to Animals and Back
SARS-CoV-2 does not only spread between humans. The virus has jumped from people into several animal species, most notably farmed mink and white-tailed deer. On mink farms in the Netherlands, genomic analysis confirmed that the virus spread from humans to mink and then back to humans, with the animal-origin strains carrying distinct genetic signatures. Researchers identified several mutations that appeared to be adaptations to mink or deer hosts, though these mutations did not appear to give the virus a meaningful advantage for human-to-human transmission.
The concern with animal reservoirs is not necessarily what is happening right now but what could happen. Every time the virus circulates extensively in a new host species, it has opportunities to acquire novel mutations. If a variant that evolved in an animal population spilled back into humans with new immune-evasion properties, it could spark a fresh wave. Surveillance of SARS-CoV-2 in animal populations has become an ongoing public health effort for exactly this reason. So far, no considerable changes to the virus’s mutation rate or evolutionary trajectory have resulted from circulation in mink and deer, but the risk is not zero and monitoring continues.