Readily achieved comparative assessment of carcinogenic risks consequent upon environmental exposures may increase understanding and contribute to cancer prevention. Procedures for hazard identification and quantitative risk assessment are established, but limited when addressing novel exposures to previously known carcinogens or any exposure to agents having only suspected carcinogenic activity. To complement other means of data evaluation, a procedure for qualitative assessment of carcinogenic risk is described. This involves categorizing the relevant carcinogen and circumstances under which exposure occurs. The categories for carcinogens are those used for hazard identification and involve whether the agent is (1) a recognized carcinogen for humans; (2) probably or (3) possibly carcinogenic for humans; (4) characterized by inadequate evidence of carcinogenicity; or (5) lacking carcinogenicity. Exposure is categorized by whether it is one which (1) establishes the agent as a recognized carcinogen; (2) is taken into account in establishing carcinogenicity status; (3) is distinct from those providing clearest evidence of carcinogenicity; (4) is not characterized in relation to carcinogenicity; or (5) involves an exposure in which absence of carcinogenic outcome is observed. These two categories of evidence allow the risk inherent in a situation to be banded as indicative of a proven, likely, inferred, unknown or unlikely carcinogenic outcome, and further characterized using sub-bands. The procedure has been applied to about fifty situations. For recognized carcinogens, including asbestos and polycyclic aromatic hydrocarbons, risks consequent upon occupational exposure, the impact of point source pollution, residence near contaminated sites and general environmental exposure are allocated across the proven band and a likely sub-band. For solvents, pesticides and other compounds having less clearly established carcinogenicity, impact on residents living near a production site, or near earlier related industrial activity is allocated to certain inferred sub-bands. Unknown carcinogenic outcome, which identifies exposure to an agent with inadequate evidence of carcinogenicity rather than being indicative of equivocal or negative data in any context, indicates both the impact of certain pollutants and user-exposure to some consumer products. Situations allocated to the unlikely risk band principally involve certain consumer products. Overall, such risk assessment may be of greatest worth in focusing community attention on proven causes of cancer and associated preventive measures.