&NA;A comparison of the prognostic values of the Dukes and Jass systems were performed with 722 patients with rectal cancer enrolled in the National Surgical Adjuvant Breast and Bowel Projects, protocol R‐01. The Jass system revealed four prognostic groups when all patients or only Dukes' B and C cases were examined; however, the magnitude of differences between groups I and II and III and IV were small. Dukes' classification, as defined in this study, revealed five prognostic groups. A statistically strong association between the Jass and Dukes systems was observed. Although histologic grade permitted further prognostic discrimination of all Dukes stages except A, only the Jass system allowed for the subdivision of C cases with up to four nodes positive for metastases. Those in that group had survival rates comparable to B cases (no nodal involvement) when scores of I and II were found. The distributions of the patients in the extremes of the Jass and Dukes systems (C2 as defined) were almost similar. The findings indicate that the Jass system is a valid prognostic method for patients with rectal carcinoma. In this material, however, it basically allowed for only two major prognostic groups whereas five were noted by the Dukes method. These results, as well as the more objective nature of Dukes' classification, warrant its continued use for prognosis and therapeutic decisions for patients with rectal cancer.