There is limited evidence for mutagenicity and carcinogenicity of trichloroethylene (TRI) in experimental test systems. Whether TRI is a human carcinogen is unclear, however. This paper presents an update and extension of a previously reported cohort of workers exposed to TRI, in total 1670 persons. Among men (n = 1421), the overall standardized mortality ratio (SMR) and cancer morbidity ratio (SIR) were close to the expected, with SMR, 0.97; 95% confidence interval (CI), 0.86 to 1.10; and SIR, 0.96; 95% CI, 0.80 to 1.16, respectively. The cancer mortality was significantly lower than expected (SMR, 0.65; 95% CI, 0.47 to 0.89), whereas an increased mortality from circulatory disorders (cardiovascular, cerebrovascular) was of borderline significance (SMR, 1.17; 95% CI, 1.00 to 1.37). No significant increase of cancer of any specific site was observed, except for a doubled incidence of nonmelanocytic skin cancer without correlation with the exposure categories. In the small female subcohort (n = 249), a nonsignificant increase of cancer and circulatory deaths was observed (SMR, 1.53 and 2.02, respectively). For both genders, however, excess risks were largely confined to groups of workers with lower exposure levels or short duration of exposure or both. It is concluded that this study provides no evidence that TRI is a human carcinogen, ie, when the exposure is as low as for this study population.