To assess the mortality experience of a cohort of chemical workers in the Kanawha Valley of West Virginia, 29,139 males who worked at any one of three facilities over a 39-year period were followed-up for vital status. The facilities include two chemical manufacturing plants and a research and development center. From this cohort, 5,785 men were found to have died as of the study end, December 31, 1978. This was less than the 6,148.5 men expected to have died, based upon the United States white male population (standardized mortality ratio (SMR) = 94, 95% confidence interval (CI) = 92-96). Eighty-six specific causes of death were examined. Statistically significant increased deaths were observed for two causes; cancers of the liver (not specified as primary or secondary) (SMR = 174; CI = 102-280) and lympho- and reticulosarcoma (SMR = 140; CI = 104-187). When all biliary and liver cancer was examined by duration and time since initial employment, the SMR for those who worked at least 25 years and whose deaths occurred 30 years or more after first employment was 301 (95% confidence limit = 168-497). The identification and follow-up of this complete cohort provides the basis for future study of subcohorts with specific chemical and process exposures and case control studies of specific causes of death.