Whether indoor air is just as polluted as outdoor air is an important question for all of the millions of people living in heavily polluted cities. Yet I’ve heard a lot of disagreement about this question. I’ve heard people tell me it’s the same — “indoor air comes from outdoors anyway, so it’s the same.” Other people have the intuition that indoor air is less polluted because it’s sealed off from the outside. Who’s right?
The Indoor Air Pollution Test
To get to the bottom of it, I tested indoor and outdoor air at 11 locations around Delhi. India is a particularly important place to test this question because air pollution averages toxic levels around India, even in the major city with India’s cleanest air.
The test was simple. At each location, I used an AirVisual Node and tested PM2.5 inside and outside (Test: is the AirVisual accurate?). For example, here’s the air pollution outside the Starbucks in Greater Kailash.
None of the places were using air purifiers. So this represents air in unpurified indoor spaces. I tested restaurants, cafes, hotels, offices, and homes. I also avoided testing in any place that kept windows or doors open constantly.
Delhi’s Indoor Air Is Safer Than Outdoor Air on Average
I calculated indoor PM2.5 pollution as a portion of outdoor pollution. For example, if indoor air is exactly as polluted as outdoor air, that percentage would be 100%. If indoor air is half as polluted, that percentage would be lower than 50%. Here’s what I found.
Nine out of 11 places had less particulate pollution than outdoor air. Only two places were exceptions.
Averaging across all sites, indoor PM2.5 in Delhi was 68% of outdoor air.
Here’s a sense of what that looks like in real life. The picture below shows PM2.5 pollution outside (left) and inside (right) a restaurant near the Saket mall. The indoor air here had 45% as much PM2.5 as air outside the restaurant.
Here’s why that’s important for people living in polluted cities: That means that simply by being indoors, we’re breathing less PM2.5 on average, even in places without air purifiers. Thus, on polluted days, I choose to work out indoors and sit indoors at cafes and restaurants rather than outdoors.
Why Do Some Places Have Cleaner Air?
When I got my first particle counter, it opened my eyes to a whole new world. All of a sudden, I could “see” all of these microscopic particles I had never seen before. After six years of testing the air, one thing I know for certain is that indoor air closely follows outdoor air.
That data from Delhi shows that same trend. When outdoor air was worse, indoor air tended to be worse too.
Of course, there is variation within that trend. One source of variation is how well sealed the rooms are. One indirect way to measure how well rooms are sealed is to look at CO2 (which the AirVisual measures). For example, at this restaurant in Delhi, CO2 was 491 PPM outdoors and 2,105 indoors.
Humans breathe out CO2. So where there are lots of humans in a well-sealed room, CO2 will rise. And because it’s better sealed, PM2.5 levels indoors will tend to be lower.
Of course, high CO2 levels can be harmful too, so there’s a tradeoff between PM2.5 pollution and CO2.
Indoor Air Pollution: Good News and Bad News
This data has good news and bad news for lungs in India. The good news is that we can breathe less particulate pollution simply by being indoors. Easy win!
The bad news is that the air pollution is still far above safe levels. Levels at all 11 places I tested were far above the World Health Organization annual limit of 10 micrograms. The “safest” place I tested was nearly 7 times the limit!
How Does Indoor Air in India Compare to Other Countries?
One interesting finding was that the benefit of being indoors was smaller in India than in China and the United States. In comparison, my tests in Shanghai found that indoor air averaged 57% of the particulate of outdoor air (results were similar for indoor air quality in Beijing). The US EPA found that American homes averaged 56% of the particulate of outdoors. At 68%, Delhi was about 10% higher.
The benefit might be smaller in India because India tends to be hot, and rooms tend to be less well sealed in places with hot climates. (However, I did these tests in January, which is fairly cold in Delhi.)
Bottom line: Indoor air in Delhi had less (68%) PM2.5 pollution than outdoor air on average. Sitting indoors is a simple way to reduce PM2.5 exposure in cities struggling with pollution across India.
Open Data and More Details
As always, I’m including all of the original data and more details on the methods for fellow data nerds.
1. Test Method
I collected data January 12th to January 29th, 2019. I collected data at each place once to avoid over-representing a single place in the data. These days were quite polluted, with PM2.5 averaging 275.8 micrograms per meter cubed. The lowest outdoor reading was 191 micrograms. The highest was 417 micrograms.
I tested air directly outside the entrance of each building, allowing 30-60 seconds for the AirVisual to give an adequately stable reading. The fact that I ran the tests in January could be relevant because rooms may be better sealed in the winter than in the summer. Because Delhi is fairly cold in January, people are more likely to close windows and doors.
The locations I tested were diverse, from a large international hotel to a regular apartment. Although there is a wide range, these places tend to be in wealthier parts of Delhi, such as Saket. If one theorizes that wealthier places tend to have better air or better-sealed rooms, then the indoor-outdoor percentage may be higher in less-well-off parts of town.
I did not test in any place that kept windows or doors open constantly. I also avoided places that were using purifiers or central air systems with filters (that I knew of). However, one exception is that the manager of the Sheraton hotel told me that the air in the lobby is filtered. At the same time, the lobby doors are opened frequently. The average in the lobby (61%) was fairly close to the overall average (68%). Thus, removing this datapoint would have minimal effect on the overall results.
3. Time of Day
The original data includes dates and times of day. Timings ranged from 8:40AM to 8:44PM, covering diverse parts of the day.
Thomas is an Associate Professor of Behavior Science at the University of Chicago Booth School of Business and the founder of, a social enterprise to help people in China breathe clean air without shelling out thousands of dollars for expensive purifiers.