This study examined the 1973 to 1989 mortality experience of Shell's two California manufacturing locations' employees who worked more than 6 months before December 31, 1989 and pensioners who were alive as of January 1, 1973. Vital status of each employee as of December 31, 1989 was determined from various sources including company records, the National Death Index, and the Social Security Administration's Master Beneficiary Record file. The study included many long-term employees, with more than half (57%) of the total population working 20 years or longer. The total population exhibited 11% lower all causes mortality and 20% lower cancer mortality, as compared with the California general population. There were no significant excesses of any cause-specific mortality including cancer. Among total employees, mortality for several cancer sites showed a statistically nonsignificant increase, for example, cancer of the kidney (8 observed deaths and 6.02 expected), cancer of the bladder (11 observed deaths and 9.17 expected), and Hodgkin's disease (2 observed deaths and 1.01 expected). A review of these work histories revealed no predominant work area or job assignment. In contrast to the ecologic studies based on local county rates, lung cancer mortality in this study was significantly lower (Standardized Mortality ratio [SMR] = 0.73). In addition, statistically significant deficits in mortality were found for cirrhosis of the liver (SMR = 0.63) and all external causes of death (SMR = 0.74). This study also failed to show an increased mortality rate for cancers of the brain, stomach, and prostate--causes which have been reported to be elevated in other refinery and petrochemical employee studies.