The cause‐specific mortality (1940–1993) of 2,985 male workers employed in three oil refineries was examined using a proportionate mortality study design. Separate analyses were undertaken by race, refinery, employment status (active and retired), and time since entry into the Oil, Chemical, and Atomic Workers (OCAW) union. Proportionate cancer mortality ratio (PCMR) analyses also were conducted. Proportionate mortality ratios (PMR) were significantly increased (P < 0.05) for cancers of the lip (PMR = 384), stomach (PMR = 142), unspecified sites of the liver (PMR = 238), pancreas (PMR = 151), connective tissues (PMR = 243), prostate (PMR = 135), eye (PMR = 407), brain (PMR = 181), benign and unspecified neoplasms (PMR = 289), and leukemia (PMR = 175) for the entire cohort. Significantly decreased mortality was observed for respiratory tuberculosis (PMR = 29), esophageal cancer (PMR = 45), rectal cancer (PMR = 49), and cancers of the bladder and other urinary organs (PMR = 40). Skin cancer was observed to be significantly increased (PMR = 242) for workers with less than 20 years since union initiation. Significantly increased PCMRs were seen for cancers of unspecified sites of the liver (PCMR = 205), brain (PCMR = 147), benign and unspecified neoplasms (PCMR = 243), and leukemia (PCMR = 146). Among nonwhites, an increased risk of bone cancer was observed in the PCMR analysis (PCMR = 704), although based on only two deaths. Analyses of mortality patterns for white males by refinery revealed similar patterns in each refinery as was seen in the overall cohort of refinery workers. Mortality patterns for whites and nonwhites also were similar. Additional analyses of deaths between 1960 and 1993 demonstrated increased mortality due to asbestosis (PMR = 683) and multiple myeloma (PMR = 124), although the multiple myeloma excess was not statistically significant. Ten deaths due to mesotheliomas were observed among these refinery workers. Am. J. Ind. Med. 33:327–340, 1998. © 1998 Wiley‐Liss, Inc.
