How We Know Masks Work

Sadly, the issue of whether or not people simply wearing a mask can help prevent the spread of COVID is a hotly debated topic. It’s no longer debated in the scientific community, but it is online and in the public. Indeed, with adults throwing produce and tantrums in the middle of supermarkets, the issues surrounding “wear-a-mask mandates” could not be a hotter topic.

How Masks Work (To Help Prevent The Spread of COVID)

The explanation for how and why they work is simple, but misinformation has confused the issue. It’s commonly believed that masks simply protect their wearer from becoming infected by filtering the air that a person inhales. While there is some evidence that cloth masks maybe offer some protection to their wearer in this regard,[1] the primary way they help mitigate the spread is by filtering the air a person exhales—by sneezing, coughing, speaking, or breathing.[2]

Why are masks effective in filtering viruses from the exhaled air? Because a virus doesn’t travel by itself; it travels in droplets of moisture propelled from the mouth that can be captured. The largest droplets are intercepted by the surface of the mask; smaller ones are stopped by impaction as they try to make it through the mask; and the smallest, including aerosols, are diffused by the mask because of something called Brownian motion–the erratic way they move.[3] This lowers the number of infected droplets in the air, and thus reduces the risk of exposure for other people.[4]

Now, it’s important to note that, while all masks filter exhalations, certain kinds of masks are better than others.[5] The early woven cloth masks were effective, but not as effective as later non-woven ones—what we today call “medical” or “surgical masks.” (According to Davies, while both types are effective, surgical masks are about three times more effective.[6]) But different kinds of cloth, layered or combined, can offer better filtration, by making the pathway particles would have to travel to get through the mask more convoluted. Aydin found that layering cotton fabrics can make homemade masks almost as effective as surgical masks,[7] and Konda found that layering different kinds of fabrics (like cotton with flannel) could also increase their efficacy.[8]  Such masks filtered 80% of particles smaller than 300 nanometers (0.3 microns) and 90% of those larger.[9] Filters can also improve mask efficacy.[10]

Now if an infected droplet is already in the air that a masked person is about to inhale, the mask will be less likely to catch that particle. That might make one question how cloth masks can filter the air that someone exhales. But that question is easily answered. Part of it is due to air flow; when you are breathing in, you are not pushing the air directly through the mask. But also–the longer a droplet is in the air, the more opportunity it has to evaporate and become smaller (and even aerosolize)—and (unless it’s smaller than 0.3 microns) the smaller it is, the less likely a particle is to be caught by a mask. So infected droplets in the air that an unmasked person has exhaled has a better chance at making it through your mask. But most of the infected droplets exiting your mouth won’t, since they have not yet had a chance to evaporate.[11] [12]

In other words, masks unquestionably work to help prevent the spread of COVID because they filter droplets (those on the right of the below picture), including most of them that would become aerosols (those in the middle), and they can even capture those that start out small (on the far left)—just a bit less efficiently. (And only about 1 in every 700 of the latter such droplets, expelled from an infected persons mouth, contain even a single virus.) [13]

So, if you are in a room with a non-masked infected person, you are pretty likely to be infected even if you are wearing a cloth mask. However, if that infected person is wearing a cloth mask, you are much less likely to be infected—even if you are not wearing a mask. This is why it is said “My mask protects you; your mask protects me.”[14] Your choice to not wear a mask does not put you at risk; it puts others at risk.

Now, since the fact that the majority of the work that masks do in curbing the spread of COVID is done by the masks that infected persons are wearing, one might wonder: why don’t we just have infected people wear them? Indeed, this was part of the reasoning behind the traditional, years long, CDC and WHO recommendations that said mask wearing is not necessary for seemingly healthy individuals. With many diseases, it is not. But then experts realized just how long a person could be infected with COVID without knowing it and even that many infected never have any symptoms.[15] Consequently, public mask mandates are the only way to ensure that infected persons—including those who are pre- or asymptomatic—are wearing a mask. Such mandates would thus reduce the number of infected people without masks, thus reduce the number of infected particles in the air, and thus reduce the probability of transmission. Once we realized this, mask wearing was recommended. Granted, the CDC and WHO lagged behind the experts in this regard, but scientists change their mind based on evidence all the time. That’s not unusual. It’s just usually not that public.

How We Know Masks Help Prevent the Spread Of COVID

So given what we know masks do, the fact that mandating them helps curb the spread of COVID is just common sense. But the evidence also bears this out.

For example, transmission rates slowed in hospitals,[16] German cities,[17] North Texas,[18] and in US states[19] after mandates went into effect. (In Germany, they reduced growth rates by 40%.) What’s more, U.S. States with mandates have seen much less spread than those without, [20] as have countries where mask use is popular.[21] Now some will argue that such correlational studies can’t prove anything because “correlation doesn’t entail causation.” But this is a misuse of that logical rule. A single correlation does not guarantee causation, but enough of them can imply it strongly enough to produce knowledge. For example, deaths rates dropped 27% after seat belt mandates were enacted in New York State,[22] and similar numbers were seen in all states after enacting such laws. Something similar happened for deaths in motorcycle accidents after helmet mandates were enacted. When the connection is obvious, and the correlation is repeated, it most definitely entails causation.

Modeling has also confirmed the effectiveness of mask mandates.[23] According to Stutt and Eikenberry, if masks are just 50% effective, they could help bring down infection rates to non-epidemic levels and reduce the death rate by as much as 45%.[24] Large reviews of observational and comparative studies have also concluded that mask mandates are highly effective,[25] as have collections of collaborating experts.[26] And, of course, there were those infected hairstylists in Missouri who didn’t pass it on to 140 of their clients because they were wearing a mask.[27] The IHME found that mask mandates could save 33,000 by October 1st,[28] and (according to Brooks) if everyone wore a mask, we could get the pandemic under control in four to eight weeks.[29] [30]

Now, if one is being stubborn, one might demand that this is not good enough. We need randomized control trials (RCTs) for mask safety and efficacy—the supposed “golden standard” in science. Well, we actually do have some. The previously mentioned study by Leung, for example, which masked some people (but not others) with respiratory illness and tested for infected droplets an aerosols found that “Surgical face masks reduced detection of coronavirus RNA in both respiratory droplets and aerosols.”[31] A lot of the other studies I mentioned above are along this line. But anti-mask activists are likely to insist that this is not good enough. We need a RTC that measures how many other people got sick with the infected wearing, or not wearing, a mask. What such arguments fail to recognize is that, while they are great (indeed necessary) for testing drugs and treatments,[32] RCTs are not necessary or even appropriate for other scientific fields or questions. Indeed, they are easily misused and can lead to confusion.[33]

To understand why, consider an example. Supposed you wanted to know about the safety and efficacy of Kevlar vests regarding their ability to protect people from bullets. So I explain the science—the physics of how Kevlar resists bullets—and show you how it works in a lab. Maybe I shoot bullets at plastic dummies, some wearing and some not wearing, Kevlar vests. I also show you correlational studies of how, say, death and injuries drop in army platoons after Kevlar vests are issued. That would be good enough right? Of course.

In fact, if anyone went further and performed a RCT on Kevlar vests, they would be pretty much useless. Why? Because a true RCT would involve lining a whole bunch of people up on a wall, giving some Kevlar vests and some not at random, and then shooting them. Obviously, that kind of study cannot ethically be done. At best—because you can’t take vests away for experimental purposes—an RCT could only compare, say, different groups of soldiers, already out in the field, who happen to be wearing vests or not (for all kinds of various reasons and in different situations). If you found a significantly lower rate of injury among those wearing Kevlar, the study would be pretty useless because it would just tell you what you already know. But if it found no significant difference in the two groups, it would still be useless. Not only is it well known that a RCT not finding a difference is not proof that it is not there, but you would immediately think that there was a random variable that skewed the results. The soldiers with the vests probably happened to find themselves in a much more dangerous situation than those without, and so it biased the study. A “failed” RCT on vests would tell you very little about the efficacy of vests.

That’s how it is with masks. We understand the science of how they work. We know they block droplets and aerosols; we know that is how COVID is spread; we know mask mandates make more people wear masks, and we know COVID is spread pre- and asymptomatically. We even have trials where their use reduces the number of infected particles in the air. Combine that with the above examples from around the world of rates dropping with mask use and mandates, and that’s all you need. We know they work. No ethically dubious RCTs, where we throw healthy people into rooms with COVID patients who may or may not be wearing masks, are necessary. And any other kind of limited RCT that we did in the field would be pretty useless. Even if it found no significant results, that wouldn’t tell us mask don’t work. Not finding something is not evidence that it is not there; random variables skewing the results would be the more likely explanation. If they are carefully designed enough, they might be able to point towards one kind of mask being more effective than the other; but proving they don’t work at all is going to be next to impossible. Field studies just can’t be controlled well enough to overturn something that is already well established. And since the evidence we already have has established that they work, we know that they do.

And that, dear reader, is how we know public masks use helps mitigate the spread of COVID-19.

[1] See Rick Kushman, “Your Mask Cuts Own Risk by 65 Percent,” UC Davis, July 6, 2020, Caitlin McCabe, “Face Masks Really Do Matter. The Scientific Evidence Is Growing,” The Wall Street Journal, July 18, 2020,  See WH Seto,et al., “Effectiveness of Precautions Against Droplets and Contact in Prevention of Nosocomial Transmission of Severe Acute Respiratory Syndrome (SARS),” The Lancet 361, no 9368 (May 3, 2003): 1519-20, and Hiroshi Nishiura, et al., “Rapid Awareness and Transmission of Severe Acute Respiratory Syndrome in Hanoi French Hospital, Vietnam,” Am J Trop Med Hyg 73, no. 1 (July 2005): 17-25, Lijie Zhang, et al., “Protection by Face Masks against Influenza A(H1N1)pdm09 Virus on Trans-Pacific Passenger Aircraft, 2009,” Emerging Infectious Diseases 19, no. 9 (September 2013): 1403-10,  See Christian J. Kähler and Rainer Hain, “Fundamental Protective Mechanisms of Face Masks Against Droplet Infections,” Journal of Aerosol Science 148 (2020) C. Raina MacIntyre, et al. “A Cluster Randomised Trial of Cloth Masks Compared with Medical Masks in Healthcare 1Workers,” BMJ Open 5 no. 4, (2015)  This study suggested that 95% of viruses in aerosols could be blocked by homemade masks, and 97% could be blocked by surgical masks: Qing-Xia Ma, et al, “Potential Utilities of Mask-Wearing and Instant Hand Hygiene for Fighting SARS-CoV-2” Journal of Medical Virology (2020) This is a study out of Hong Kong which suggested that people wearing a mask was very effective at reducing transmission of alpha coronaviruses”; Nancy H. L. Leung, et al., “Respiratory Virus Shedding in Exhaled Breath and Efficacy of Face Masks,” Nature Medicine 26 (2020): 676-80,

[2] See also WH Seto, et al., “Effectiveness of Precautions Against Droplets and Contact in Prevention of Nosocomial Transmission of Severe Acute Respiratory Syndrome (SARS),” The Lancet 361, no 9368 (May 3, 2003): 1519-20,; see also Hiroshi Nishiura, et al., “Rapid Awareness and Transmission of Severe Acute Respiratory Syndrome in Hanoi French Hospital, Vietnam,” Am J Trop Med Hyg 73, no. 1 (July 2005): 17-25; Harvey Fineberg, Rapid Expert Consultation on the Possibility of Bioaerosol Spread of SARS-CoV-2 for the COVID-19 Pandemic (April 1, 2020), (The National Academies Press, 2020), chapter 1 and 2,

[3] Depending on your definition, aerosols range from around 100 microns to 0.1 micron. “Various sources will put the cutoff at 2 µm, 5 µm, 10 µm, 20 µm, or even 100 µm.” Justin Morgenstern, “Aerosols, Droplets, and Airborne Spread: Everything You Could Possibly Want to Know,” First10EM, April 6, 2020, For simplicity, I’ll define aerosol as a droplet that is 10 µm in size. Brownian motion dominates in particles less than 0.3 µm in size. 

[4] Talib Dbouk and Dimitris Drikakis, “On Respiratory Droplets and Face Masks,” Physics of Fluids 32, no. 063303, published electronically June 16, 2020, Bhanu Bhakta Neupane, Sangita Mainali, Amita Sharma, and Basant Giri, “Optical Microscopic Study of Surface Morphology and Filtering Efficiency of Face Masks,” PeerJ 7,no. e7142 (2019),

[5] Cloth masks of only one material seem to have very little effectiveness: Samy Rengasamy, Benjamin Eimer, and Ronald E. Shaffer, “Simple Respiratory Protection – Evaluation of the Filtration Performance of Cloth Masks and Common Fabric Materials Against 20-1000 nm Size Particles,” Annals of Occupational Hygiene 54, no. 7 (October 2010): 789-98, This is why those who are just wearing bandanas or pulling their t-shirt over their mouth, are not doing anyone much good.

[6]  Both types “significantly reduced the number of microorganisms expelled by volunteers,” “the surgical mask was 3 times more effective.”) Anna Davies, et al., “Testing the Efficacy of Homemade Masks: Would They Protect in an Influenza Pandemic?” Disaster Medicine and Public Health Preparedness 7, no. 4 (August 2013): 413-8,;  Milton (2013) found that surgical masks decreased emission of large particles by 25 fold, and aerosols by 3 fold in flu patients. See Donald K Milton, et al., Influenza Virus Aerosols in Human Exhaled Breath: Particle Size, Culturability, and Effect of Surgical Masks,” PLoS Pathogens 9, no. 3 (March 2013): 1-7,

[7] Aydin et al. (2020), suggests that layering greatly increases the filtering efficiency of cloth masks while also maintaining some breathability, Onur Aydin, et al., “Performance of Fabrics for Home-Made Masks Against the Spread of Respiratory Infections through Droplets: A Quantitative Mechanistic Study,” medRxiv,preprint, submitted July 8, 2020

[8] Abhiteja Konda, Abhinav Prakash, Gregory A. Moss, Michael Schmoldt, Gregory D. Grant, and Supratik Guha, “Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks,” ACS Nano 14, no. 5 (2020): 6339-47,

[9] “Overall, we find that combinations of various commonly available fabrics used in cloth masks can potentially provide significant protection against the transmission of aerosol particles.” Konda, “Aerosol Filtration Efficiency of Common Fabrics Used in Respiratory Cloth Masks.”

[10] Dr. Marty, a professor of infectious diseases at Florida International University told Good Morning America “But if you add that filter, then you’re also adding a really good protection for yourself.” See Becky Worley, Anthon Kane, Robyn Weil, and Angeline Jane Bernabe, “Face Masks With Filter add Another Layer of Protection, Experts Say,” GMA, July 16, 2020,

[11] This is actually what N95 masks are named for; they protect their wearer by being 95% effective at filtering the air that a person breathes in. And because they often have unfiltered exhaust vents, they usually don’t filter the air a person breathes out. Combine that with the fact that they don’t even perform their intended function well unless they are perfectly fit, and you can realize why they should likely only be worn by health care workers in high risk environments.

[12] Some clarification here is useful. Technically, depending on how you classify “aerosols” (definitions range from 5 microns to 100 microns), most of  the particles you breath out could be classified as aerosols–and depending on their size, the mask will filter them with different efficiencies. Even cloth masks are very good at filtering down to 10 microns, pretty good down to 5 microns, but not great below 5. Neupane, “Optical Microscopic Study of Surface Morphology and Filtering Efficiency of Face Masks.”

In one way, this is concerning because according to Burch (2020), the average size for aerosols leaving your mouth is 3 microns.  The good news is, despite the fact that they make up the largest number, they only represent 0.00024% of the liquid leaving your mouth during a cough. Consequently, very few of them are infected (at worst 1 out of every 700). A full 99.99976% of the viruses sprayed during a cough are carried in droplets — not aerosols.’) So the majority of transmission happens from droplets.  What’s more, the deadliest aerosols are those that started out as droplets, but then evaporated down; they have higher concentrations of the virus. Masks catch those. So the inability of masks to filter out 0.3 micron particles does not greatly hinder their ability to keep infected particles out of the air, and thus does not prevent them from efficiently preventing the spread of COVID.
                Adrien Burch, “A Microscopic Perspective on Airborne COVID-19,” The Medium, March 31, 2020,

 [13] Adrien Burch, “A Microscopic Perspective on Airborne COVID-19,” Medium, March 31, 2020,

[14] From the abstract of: Richard O. J. H. Stutt, Renata Retkute, Michael Bradley, Christopher A. Gilligan, and John Colvin, “A Modelling Framework to Assess the Likely Effectiveness of Facemasks in Combination with ‘Lock-down’ in Managing the COVID-19 Pandemic,” Proceedings of the Royal Society A (2020)

[15] Wycliffe E Wei, Zongbin Li, Calvin J Chiew, Sarah E Yong, Matthias P Toh, and Vernon J Lee, ”Presymptomatic Transmission of SARS-CoV-2-Singapore, January 23-March 16, 2020,” MMWR Morb Mortal Wkly Rep. 69, no. 14 (April 2020): 411-5,

 [16]Anthony D Sung, et al. “Universal Mask Usage for Reduction of Respiratory Viral Infections after Stem Cell Transplant: a Prospective Trial,” Clin Infect Dis 63, no. 8 (October 2016): 999-1006, Xiaowen Wang, Enrico G. Ferro, Guohai Zhou, et al., “Association Between universal Masking in a Health Care System and SARS-CoV-2 Positivity Among Health Care Workers,” JAMA,published electronically July 14, 2020, 10.1001/jama.2020.12897. In this study, cases of COVID-19 declined after mask mandates were put into effect in hospitals (that required all health care workers and patients to mask up). The study concluded that such mandates reduce the transmission of SARS-CoV-2.

[17] Timo Mitze, Reinhold Kosfeld, Johannes Rode, and Klaus Wälde, “Face Masks Considerably Reduce COVID-19 Cases in Germany: A Synthetic Control Method Approach,” IZA (June 2020)

This study shows the impact of mask mandates in Germany. In Jena, for example, the first German city to enact such a mandate, COVID-19 cases fell by almost 25% in 20 days. The study concluded that similar mandates could ruse the daily growth rate by 40% in the long term, although it did acknowledge that, outside Germany, different norms and climatic conditions in other countries might result in different protective outcomes.

[18] “HSC COVID-19 Report #5 – July 20, 2020,” University of North Texas Health Science Center at Fort Worth,

[19] In those 15 US States, they likely prevented up to 450,000 cases in under two months. Wei Lyu and George L. Wehby, “Community Use of Face Masks and COVID-19: Evidence from a Natural Experiment of State Mandates in the US,” Health Affair 39, no. 8 (2020): 1-7, 10.1377/hlthaff.2020.00818.

This was a retrospective analysis which examined the effects that different governmental orders to wear face masks had on COVID-19 growth rates, from April 9-May 15, 2020. It estimated that they prevented between 230,000 and 450,000 cases by May 22 (a reduction of 14-27%).

[20] Kasra Zarei and John Duchneskie, “Coronavirus Cases Rise in States with Relaxed Face Mask Policies,” The Philadelphia Inquirer, June 24, 2020,

[21] American Thoracic Society, “Countries with Early Adoption of Face Masks Showed Modest COVID-19 Infection Rates,” Medical Xpress, June 24, 2020,

[22] Joseph Berger, “Death Drops 27% With State’s Seat-belt Law, The New York Times, May 1, 1985,

[23] Samantha M. Tracht, Sara Y. Del Valle, and James M. Hyman, “Mathematical Modeling of the Effectiveness of Facemasks in Reducing the Spread of Novel Influenza A (H1N1),” Plos One 5, no. 2 (February 2010): 1-12,

[24] Stutt, et al., “A Modelling Framework to Assess the Likely Effectiveness of Facemasks in Combination with ‘Lock-down’ in Managing the COVID-19 Pandemic.”

To keep the infection rate (R0) below 1.0, the authors argue for widespread use of face masks. “[F]acemask adoption by entire populations would have a significant impact on reducing COVID-19 spread.” “[I]n summary, our modelling analyses provide support for the immediate, universal adoption of facemasks by the public.”

[25] Derek K Chu, et al., “Physical Distancing, Face Masks, and Eye Protection to Prevent Person-to-person Transmission of SARS-CoV-2 and COVID-19: A Systematic Review and Meta-analysis,” The Lancet 395, no. 10242 (2020): 1973-87,

This was a review of 172 observational studies and 44 relevant comparative studies. The authors concluded “Face mask use could result in a large reduction in risk of infection.”

[26] Kimberly A. Prather, Chia C. Wang, and Robert T. Schooley, “Reducing Transmission of SARS-CoV-2,” Science 368, no. 6498 (June 2020): 1422-24, In this paper, aerosol chemists and an infectious disease specialist argue that, because “airborne spread from undiagnosed infections will continuously undermine the effectiveness of even the most vigorous testing, tracing and social distancing programs,” the widespread use of masks are necessary to help prevent the spread of COVID. Both analytical information about the virus and information about countries where masks are commonplace was used.

Catherine M. Clase, et al., “Cloth Masks May Prevent Transmission of COVID-19: An Evidence-Based, Risk-Based Approach,” Annals of Internal Medicine, published electronically May 22, 2020, This study, done by an international research team of medical doctors and other medical specialists not only concluded that cloth masks worn by the public will reduce COVID-19 transmission rates but that the benefits of widespread mask use outweigh any risks that may be brought about by wearing masks (such as improper use).

 [27] M. Joshua Hendrix, Charles Walde, Kendra Findley, and Robin Trotman, “Absence of Apparent Transmission of SARS-CoV-2 from Two Stylists After Exposure at a Hair Salon with a Universal Face Covering Policy – Springfield, Missouri, May 2020,” Weekly 69, no. 28 (July 1, 2020): 930-32,

[28] “New IHME COVID-19 Model Projects Nearly 180,000 US Deaths,” IHME,June 24, 2020,

[29] For the quote, see McCabe, “Face Masks Really Do Matter.” For the evidence behind it, see John T. Brooks, Jay C. Butler, Robert R. Redfield, “Universal Masking to Prevent SARS-CoV-2 Transmission – The Time is Now,” Jama, published online July 14, 2020,

[30] For more such evidence, see “Face Masks – A Summary of Relevant Research Papers for COVID-19,” Sound Reason & More,June 11, 2020,

[31] Leung et al. 2020. Respiratory virus shedding in exhaled breath and efficacy of

face masks. Under review. DOI: 10.21203/

[32] Markus MacGill, “What is a Randomized Controlled Trial?” Medical News Today, December 4, 2018,

 [33] Rebecca A. Clay, “More than One Way to Measure,” American Psychological Association 41, no. 8 (September 2010): 52,; Roger Mulder, et al., “The Limitations of Using Randomised Controlled Trials as a Basis for Developing Treatment Guidelines,” Evidence-Based Mental Health 21, no. 1 (2018): 4-6,


3 thoughts on “How We Know Masks Work

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s