To evaluate the possible association between occupational exposure to benzene and subsequent death from leukemia, the National Institute for Occupational Safety and Health (NIOSH) conducted a retrospective cohort mortality study of workers who had been exposed to benzene in the manufacture of rubber hydrochloride at two locations in Ohio. Ascertainment of vital status was accomplished for 98% of the cohort. Among 748 workers who had at least one day of exposure to benzene between 1940 and 1950, seven deaths from leukemia occurred; from United States death rates standardized for sex, age, and calendar time period, only 1.25 leukemia deaths would have been expected (standardized mortality ratio = 560; p less than 0.001). Mean duration of exposure to benzene was brief, and 437 (58%) of the cohort were exposed for less than 1 year. Evaluation of leukemia mortality for those workers exposed five or more years showed an SMR of 2100. All leukemia deaths were myelocytic or monocytic in cell type. Four additional cases of leukemia have been reorganized in workers at the study locations, but occurred in persons not encompassed by the strict definition of the cohort. Reconstruction of past exposures to benzene at the two locations indicates that in some areas of the plant airborne benzene concentrations rose occasionally to several hundred parts per million (ppm), but that for the most part, employee eight-hour time-weighted averages (TWA) fell within the limits considered permissible at the time of exposure. These data corroborate an initial analysis of the same cohort by Infante et al, and indicate that benzene is a human carcinogen at a range of exposures not greatly above the current legal standard.