Cancer risk from toxic air pollution, every county, 1996 to 2021
EPA estimated lifetime cancer risk from breathing outdoor air toxics in every census tract, starting in 1996. In 2026 it released the 2021 pollution concentrations but not the risk numbers. The 2021 map here applies EPA's own method to that data. Counties are population-weighted averages of tracts. These are screening estimates, not measured cases.
Over time
This countyU.S. average
Click a row to open it on the map.
#
County
Risk per million
vs U.S.
How to read this
Risk per million = EPA's estimate of extra cancers per million people if they breathed that air for 70 years. EPA's assessment is meant to find places and pollutants worth a closer look. It is not a health study.
Years are not a clean trend. EPA changed its emissions inventory, models and toxicity values between assessments, and says so. The "rescored" view puts every year on the 2020 toxicity values, which removes one of those changes (the big ones: formaldehyde before 2005 and ethylene oxide before 2014). Emissions and models still differ. Metal and PAH groups are not rescored.
2021 is ours. EPA published 2021 block-level concentrations; we multiplied them by EPA's 2020 unit risk values (71 of 72 pollutants reproduce EPA's 2020 numbers exactly). One pollutant group, PAH/POM (about 4% of risk), is estimated. 2014-2021 already share one toxicity table, so rescoring changes nothing there.
EPA rounded tract totals to one significant figure for 2017-2019, so county averages in those years cluster at 10, 20 and 30.
County boundaries are 2020. Older years use their own tract IDs, grouped by county code (a few renamed counties remapped). U.S. Virgin Islands are in the data but not drawn.