In response to reports of an unusually high number of colorectal cancers among employees on a unit devoted to the manufacture of polypropylene, we examined colorectal cancer incidence rates for 335 workers with at least 6 months employment on this unit from 1960 to 1985. Assuming a 10-year latent period, we found a significant 5.6-fold colorectal cancer excess (7 observed/1.26 expected, standardized incidence ratio (SIR) = 5.6, 95% confidence interval (CI) 2.2 to 11.5), concentrated among mechanical (5 observed/0.47 expected, SIR = 10.6, 95% CI 3.4 to 24.7) and process workers (2 observed/0.40 expected, SIR = 5.0, 95% CI 0.6 to 17.8). No colorectal cancer was found among administrative/office personnel on our study roster. All of the cancers occurred in employees who initially worked on the unit during its first 6 years of operation, and the minimum interval from start of employment to cancer development was 20 years. In addition, the ages of the cancer cases suggested a younger age distribution than would be expected based on general population rates. Since this study was descriptive in nature, we could not evaluate whether the cancer excess was related to occupational exposures, other environmental factors, or the random clustering of cancers in our worker populations (viz, "chance"). Further studies are currently ongoing to assess occupational and/or personal factors that may be related to this colorectal cancer excess.