The mortality experience of workers at seven beryllium processing facilities in the US was analyzed in a retrospective study. The cohort consisted of 9225 male workers who had worked for at least 2 days between January 1, 1940 and December 31, 1969. Vital status was determined from Social Security Administration, Internal Revenue Service, and post office records. Life table analyses were conducted. The influence of geographic variation in lung cancer mortality was evaluated using county lung cancer rates. Adjustment for smoking habits was based on a 1968 survey of 1466 members of the cohort. Results showed that 5681 men (61.6%) were still alive, 3240 (35.1%) were deceased, and 304 (3.3%) were of unknown vital status. The standardized mortality ratio (SMR) for lung cancer in the entire cohort was 1.26. Significant SMRs were evident for two of the oldest facilities in Lorain, Ohio, and Reading, Pennsylvania. In addition, significantly elevated SMRs were found for the categories all deaths, ischemic heart disease, pneumoconiosis and other respiratory diseases, and chronic and unspecified nephritis, renal failure, and other renal sclerosis. While lung cancer SMRs did not significantly increase with longer employment durations, they did increase with longer time since first exposure. The SMR for lung cancer was particularly elevated (3.33) for workers with history of acute beryllium disease at the Lorain facility. Elevated lung cancer SMRs in four of five facilities operating in the 1950s showed that the mandated controls of the 1940s had little effect. Since neither smoking nor geographic location adequately explained the increased lung cancer risks, the authors conclude that exposure to beryllium (7440417) was the most likely reason for the increased lung cancer SMRs observed in the oldest facilities in this study.