Results from a prospective mortality surveillance of 3803 refinery and petrochemical workers at a Shell Oil Company facility in Louisiana are presented. This report includes employees who worked more than 6 months before January 1, 1994 and pensioners who were alive as of January 1, 1973. Vital status was ascertained through 1993. Regardless of the comparison population used to calculate expected numbers (United States, Louisiana, or the surrounding tri-parish area), significantly fewer deaths were observed for all causes combined, all malignant neoplasms, heart disease, nonmalignant respiratory disease, and cirrhosis of the liver among male employees after 10 or more years' latency. With the United States as comparison, the all causes combined standardized mortality ratio (SMR) was 0.72 (95% confidence interval [CI] = 0.65 to 0.79), and the SMR for all cancer was 0.75 (95% CI = 0.61 to 0.92). The brain cancer rate for this group was nonsignificantly increased, with five observed deaths and three expected deaths, whereas mortality from leukemia was consistently lower than expected. The overall favorable mortality experienced by employees at this refinery and chemical plant is probably a result of a combination of factors, such as the healthy worker effect, relatively low risks related to the workplace, and the beneficial effects of continuing employment.