A cohort study is presented on the mortality of blue-collar workers in an asbestos-cement production plant that has been operating since 1907. Use of both crocidolite and chrysotile is reported. The cohort includes 2608 men and 759 women who were employed in the plant on Jan. 1, 1950 and those who started to work between 1950 and 1980. Follow-up was terminated on April 15, 1986 with 97.9% traced. Expected deaths were estimated from the age- and sex-specific regional mortality rates for the years 1969 to 1981. The data have been analyzed for the period 1964 to 1986 based on person-years at risk: 43,000 for men and 14,494 for women. A statistically significant increase was found in both sexes for mortality from all causes. From 1964 to 1986, 728 men died from all causes (608 expected), 275 with cancer at any site (158 expected) 110 with lung cancer (41 expected), 28 with pleural tumors (1 expected) and 85 with asbestosis (less than 1 expected). Corresponding figures for women were--all causes: 136 deaths versus 102 expected; all cancers: 79 verses 32 expected; lung cancer: 7 versus 2 expected; pleural tumor: 15 versus 0 expected and asbestosis: 4 versus 0 expected. Deaths from digestive tract cancer were in excess only among women (18 observed versus 10 expected, p less than 0.01). No excess was found for deaths from laryngeal cancer. Standardized mortality rates (SMR) for lung cancer among males showed a clear increase in direct relationship with length of follow-up. SMR according to length of employment were 234 for length 10 to 19 years, 363 for 20 to 29 years, and 256 for 30 years or longer (p less than 0.05 and lower).