Several studies of styrene-butadiene rubber (SBR) workers have reported excess cancers at various sites; however, little could be concluded concerning specific etiologic agents because of the multiple exposures encountered by these groups. The current study examined cause-specific mortality in a cohort of 2,586 male workers employed for at least 6 months between 1943 and 1979 in a butadiene manufacturing plant that supplied butadiene to two SBR plants. Standardized mortality ratios were calculated using national (NSMR) and local (LSMR) comparison populations. The all-cause NSMR was 80 (p less than 0.05) and the all-cancer NSMR was 84; the corresponding LSMRs were 96 and 76 (p less than 0.05). No significant excesses were observed for any cause of death except lymphosarcoma and reticulum cell sarcoma (NSMR = 235). When the cohort was subdivided into routine, nonroutine, and low-exposure groups, the SMRs were consistently elevated for this cause of death in all three groups. However, direct comparisons between each of the two exposure groups and the low-exposure group were inconsistent. This suggests butadiene may not be responsible for the excess, but the association deserves close attention in future studies.