An update of a cohort study of 4855 employees at a Paulsboro, New Jersey refinery was conducted to further examine mortality patterns. The earlier study investigated refinery workers employed for a minimum of 1 year between 1 January 1946 and 1 January 1979. The vital status of these workers was ascertained through 1979. The update extended enrollment in the study and vital status follow-up for an additional 8 years (1980-1987). As in the previous study, mortality from all causes [standardized mortality ratio (SMR) = 87; 95% confidence interval (95% CI): 83-91] was significantly lower than expected compared with the general population. Total cancer mortality was also lower than expected (SMR = 96; 95% CI: 86-106). A borderline significant mortality increase in prostatic cancer was found (SMR = 144; 95% CI: 106-190). This increase was similar to the nonsignificant increase reported in the original study (SMR = 135; 95% CI: 90-196). The excess was of comparable magnitude among white males and nonwhite males, although it was not significant for the latter. Detailed analysis indicated that the prostatic cancer was not likely to be related to employment at the refinery. Mortality from lymphatic and hematopoietic cancers was similar to the expected mortality. Mortality from overall leukemia was as expected and detailed analyses by specific cell type showed no increase. An increase in mortality occurred from non-Hodgkin's lymphoma among male workers (SMR = 132; 95% CI: 74-217). The increase was not statistically significant and unlikely to be associated with refinery employment. Mortality from multiple myeloma among male employees was lower than expected (SMR = 74; 95% CI: 20-190). Mortality from asbestos-related diseases (pulmonary fibrosis, lung cancer, malignant mesothelioma) was also lower than expected among male workers. No cause-specific mortality was found to be associated with duration of employment at the refinery, including several causes which have been reported to be elevated in previous studies. The findings of this updated study indicate, as in the previous report, the generally favorable mortality experience of Paulsboro refinery workers.