Objective-A retrospective cohort study of 640 male polypropylene production workers in Germany was performed to evaluate the reported association between colorectal cancer and polypropylene. Method-The follow up period was 1956 to 1990. Expected numbers of cancers were derived from incidence rates adjusted for age and calendar year from the Saarland cancer registry." Results-Three colorectal cancers were identified compared with 4-0 expected (standardised incidence ratio (SIR) = 0 75, 95% confidence interval (95% CI): 0.15-2.19). For total cancers there were 27 cases in the cohort compared with 35*4 expected (SIR = 0-76, 95% CI: 0-50-1.11).Discussion-These results do not support earlier reports of a link between polypropylene production and colorectal cancer, but are consistent with a number of recent investigations of polypropylene production workers that have reported no association with risk of colorectal cancer. Due to the small size of this and other similar studies, however, a small to moderate increase in risk cannot be ruled out.