A risk assessment using quantitative methods for the measurement of cancer risks from the occupational exposure to ethylene-dibromide (106934) (EDB) was carried out at the request of NIOSH. This involved an examination of the potential carcinogenicity of EDB to humans over a lifetime at five exposure levels. The rationale for quantitative risk assessment included discussions on epidemiologic investigations, cancer bioassays, carcinogenic thresholds, shapes of dose response curves at low doses, interspecies conversion, overall sensitivity to carcinogenesis, scaling factors, benign versus malignant tumors, other influencing factors, and short term assays. Results of four rodent assays for the carcinogenicity of EDB were reviewed. Skin painting and gavage appeared to be inappropriate, while two inhalation studies were considered suitable. The rationale for the combined use of data points from the two studies was presented. The experimental results were conceptualized as several mathematical models (probit, logit, one-hit, gamma multihit, Armitage/Doll Multistage, and statistico/pharmacokinetic) which were described. Statistical methods were outlined. Of the models described, the linearized multistage and gamma multihit models were used and fitted using the GLOBAL computer program. Results showed that, at all the designated ambient occupational concentrations of EDB, the lifetime cancer risk was quite high, particularly at 10 parts per million (ppm) and 20ppm, the current federal occupational standard for an 8 hour time weighted average. Many industrial hygiene surveys have shown that typical occupational exposures range from nondetectable to about 1ppm, with some environments, such as truck trailers loaded with fumigated produce, having up to 20ppm. The authors conclude that, even at the typical exposure levels, there may be a risk of cancer.