Excessive leukemia mortality has appeared consistently in epidemiological studies of British and U.S. rubber industry workers. Attempts to identify causative factors have focused on exposure to benzene and other solvents. Interpretations of findings from these studies have often been influenced by expectations of a benzene/nonlymphocytic leukemia association, seen from previous work in other settings. However, data from the rubber industry studies have not been consistent with this expectation, as lymphocytic and nonlymphocytic leukemia have shown similar mortality excesses. Data from a small case-control study of lymphocytic leukemia are presented to illustrate an approach that considers multiple solvent exposures. The associations with lymphocytic leukemia risk observed for a number of solvents, most notably carbon tetrachloride and carbon disulfide, were stronger than those detected for benzene.