This retrospective follow-up study evaluated the mortality experience of 4,323 men employed at a plant in Alabama (AL) that manufactures agricultural and other chemicals. On average, there were 18 years of follow-up per subject during the study period of 1951 to 1987. The observed numbers of deaths among cohort members were compared with the numbers expected on the basis of United States (US) and AL general population mortality rates. The all causes standardized mortality ratio (SMR), computed using US rates as the referent, was 97 (233 observed/240 expected deaths) for whites and 68 (47/69) for blacks. White subjects had more than expected deaths from buccal cavity and pharynx (BCP) cancer [SMR = 388; 95% confidence interval (CI) = 125–905] and from esophageal cancer (SMR = 417; 112–1,067). Their lung cancer mortality rate was 50% higher than the rate of US white men and 14% higher than the rate of AL white men. Each of these three cancers has strong nonoccupational determinants, the roles of which were not assessed and which may have been responsible in whole or in part for the observed increases. The excesses of lung and esophageal cancer were concentrated among short-term employees, an observation which also argues against a causal link with occupational factors. Black men experienced no increased mortality from BCP, esophageal or lung cancers, but results for blacks were imprecise. For white and black subjects combined, there were 3 observed versus 0.62 expected deaths due to soft tissue sarcoma (p = 0.05). The job histories of subjects with this type of cancer did not suggest any shared occupational exposure.