The mortality experience of 594 individuals occupationally exposed to benzene was investigated using a retrospective cohort design. Three hundred thirty-five of the employees began working in benzene areas prior to 1950, which provided a sound data base from which to examine latency. Data derived from work histories and industrial hygiene records permitted estimation of exposure intensities and cumulative dosages for each employee. No mortalities directly attributable to benzene exposure were observed. Several cases of leukemia and other blood disorders were noted and discussed.