An estimation of the human lung cancer 'unit risk' from diesel engine particulate emissions has been made using a comparative potency approach. This approach involves evaluating the tumorigenic and mutagenic potencies of the particlates from four diesel and one gasoline engine in relation to other combustion and pyrolysis products (coke oven, roofing tar, and cigarette smoke) that cause lung cancer in humans. The unit cancer risk is predicated on the linear nonthreshold extrapolation model and is the individual lifetime excess lung cancer risk from continuous exposure to 1 microgram carcinogen per cu m inhaled air. The human lung cancer unit risks obtained from the epidemiologic data for coke oven workers, roofing tar applicators, and cigarette smokers. The comparative potencies of these three materials and the diesel and gasoline engine exhaust particulates (as organic extracts) were evaluated by in vivo tumorigenicity bioassays that proved suitable for this analysis: Ames Salmonella microsome bioassay, L5178y mouse lymphoma cell mutagenesis bioassay, and sister chromatid exchange bioassay in Chinese hamster ovary cells.