Jump to main content.


1996 National Air Toxics Assessment Exposure and Risk Data

Information provided for informational purposes onlyNote: EPA no longer updates this information, but it may be useful as a reference or resource.

IMPORTANT: These data are provided for those who wish to do their own technical analyses and comparisons using the most refined output available from this national-scale assessment. In performing such analyses, it is extremely important that users be mindful of the purposes for which the national-scale assessment was developed. In our judgment, it would be inappropriate to use these results alone to draw conclusions about local concentrations and risk. The results are most meaningful when viewed at the State or national level. There are important limitations that affect how the data should be used and interpreted. These limitations are related to gaps in emissions data, limitations in the computer models used, default assumptions in the risk assessment and limitations in the overall design of the assessment (intended to address some questions but not others). Please see the limitations section of this website for more details. All data are in Microsoft Excel spreadsheets that have been zipped. Even zipped, the files are very large and may have extended download times.

Individual Pollutant Results of the Exposure Characterization at the Census Tract Level

Pollutant CAS # SOROAD # File Download
1,3-Butadiene 106990 43218
Formaldehyde 50000 43502
Acetaldehyde 75070 43503
Acrolein 107028 43505
Ethylene Oxide 75218 43601
Acrylonitrile 107131 43704
Methylene Chloride 75092 43802
Chloroform 67663 43803
Carbon Tetrachloride 56235 43804
Ethylene Dichloride 707062 43815
Perchloroethylene 127184 43817
Tricholorethylene 79016 43824
Ethylene Dibromide 106934 43837
Propylene Dichloride 78875 43838
Vinyl Chloride 75014 43860
Benzene 71432 45201
1,3-Dichloropropene 542756 80152
Hexachlorobenzene 118741 80183
Hydrazine 302012 80188
POM(total) N/A 80230
PCBS 1336363 80231
Quinoline 91225 80239
1,1,2,2,-TCA 79345 80246
Arsenic Compounds N/A 82103
Beryllium Compounds N/A 82105
Cadmium Compounds N/A 82110
Chromium Compounds N/A 82112
Lead Compounds N/A 82128
Manganese Compounds N/A 82132
Nickel Compounds N/A 82136
Mercury Compounds N/A 82142
Coke Oven Emissions N/A 80411
Diesel PM N/A 80400

Aggregate Results of the Risk Characterization at the Census Tract Level

Cancer Risk
Non-Cancer Respiratory Irritation Risk

Individual Pollutant Results of the Exposure Characterization at the County/State Level

Pollutant CAS # SOROAD # File Download
1,3-Butadiene 106990 43218
Formaldehyde 50000 43502
Acetaldehyde 75070 43503
Acrolein 107028 43505
Ethylene Oxide 75218 43601
Acrylonitrile 107131 43704
Methylene Chloride 75092 43802
Chloroform 67663 43803
Carbon Tetrachloride 56235 43804
Ethylene Dichloride 707062 43815
Perchloroethylene 127184 43817
Tricholorethylene 79016 43824
Ethylene Dibromide 106934 43837
Propylene Dichloride 78875 43838
Vinyl Chloride 75014 43860
Benzene 71432 45201
1,3-Dichloropropene 542756 80152
Hexachlorobenzene 118741 80183
Hydrazine 302012 80188
POM(total) N/A 80230
PCBS 1336363 80231
Quinoline 91225 80239
1,1,2,2,-TCA 79345 80246
Arsenic Compounds N/A 82103
Beryllium Compounds N/A 82105
Cadmium Compounds N/A 82110
Chromium Compounds N/A 82112
Lead Compounds N/A 82128
Manganese Compounds N/A 82132
Nickel Compounds N/A 82136
Mercury Compounds N/A 82142
Coke Oven Emissions N/A 80411
Diesel PM N/A 80400

Aggregate Results of the Risk Characterization at the County/State Level

Cancer Risk
Non-Cancer Risk

Q: Why don't the nationwide cumulative cancer risk estimates in the table of county risk levels match the maps on this website? The table on this page shows a median nationwide upper bound cancer risk levels of about 50 in a million, but the map indicates the 50th percentile cumulative risk level is only about 30 in a million.

A: In the table, the percentiles shown for each county reflect the risks of all the census tracts in the county, and the nationwide percentiles at the top of the table reflect all the census tracts in the US -- not the county medians. In contrast, the map legend reflects the percentiles for the county medians, not the individual tracts. The reason the nationwide median of county risks is different from the nationwide median for tract risks is that different counties contain vastly different numbers of census tracts. Furthermore, counties with larger numbers of tracts tend to be urban (and therefore higher-risk). All these high-risk tracts get represented on the map only by a single county median. In the other extreme, some rural counties have only a one census tract, and that single tract's risk therefore represents an entire county. For this reason, the median of county risks is much more influenced by rural census tracts -- and is correspondingly lower -- than the median of census tract risks.


Local Navigation


Jump to main content.