In order to test the hypothesis that leukemia and lymphoma may be associated with exposure to styrene, cause-of-death patterns were studied at two reinforced plastic boatbuilding facilities. There were 5,021 workers who met the criterion of having worked at the two plants between 1959 and 1978. Based on industrial hygiene surveys conducted at the two plants, 2,060 individuals were determined to have worked in departments classified as having high exposure to styrene. There were 176 deaths observed among the total cohort, in comparison to 195.3 deaths expected (standardized mortality ratio [SMR] = 90). Among the high styrene exposure group, 47 deaths were observed in comparison to 41.5 deaths expected (SMR = 113). No leukemia or lymphoma deaths were observed in either the high exposure group or the total cohort. An excess of accidental deaths among the high exposure group (18 observed vs 12.4 expected) did not appear to be due to styrene's acute central nervous system effects, since none of the individuals were employed at the facilities at the time of their death. However, the possibility of a chronic effect could not be dismissed. While leukemia and lymphoma did not appear to be related to styrene exposure, the study had little statistical power to detect excesses of these diseases, due to the short length of observation and the young age of most cohort members. It was concluded that the cohort should provide increasingly useful information on chronic styrene toxicity as the population ages and the observation period increases.