The authors investigated the relationship between flow cytometric DNA index (DI, defined as the ratio of the DNA content of malignant cells to that of normal cells) and other prognostic factors (grade and stage, anatomical site, age and sex) with the survival of 115 patients with colorectal cancer. Multiple biopsy specimens from 62 patients were taken during colonoscopy before surgery. Additional samples from 53 patients were obtained from paraffin-embedded material. All patients were treated with surgery only. Fresh-frozen material gave higher incidence of DNA aneuploidy than paraffin-embedded material (79% versus 41%). The patients with DNA diploid tumors (DI = 1) had a better overall survival than those with DNA aneuploid tumors (DI = 1). Among DNA aneuploid tumors, those with DI greater than 1.2 (excluding DI = 2) were worse than those with DI = 1.2 (excluding DI = 1) and DI = 2. Cox's regression analysis showed that pathologic stage was more important for prognosis than DNA index, whereas age, sex, histologic grade, and anatomic site were removed from the analysis as not relevant for prognosis. Relative risk of death (RR), in reference to patients with DI = 1 and Stages A + B (RR = 1), were RR = 1.8 for patients with carcinomas with Stage C. RR = 2.7 for patients with carcinomas with DNA near-diploid and DNA tetraploid tumors. RR = 3.5 for those with DI greater than 1.2 (excluding DI = 2), and RR = 8.0 for those with Stage D. These data indicate that flow cytometrically evaluated DI values have a relevant independent power for predicting the clinical outcome of colorectal cancer patients.