The purpose of this study was to examine the influence of race, gender, and age on colorectal cancer cases in our tumor registry between January 1987 and December 2000 and to determine the implications of these factors on screening strategies. Tumors were defined as early (Stage I/II) or late (Stage III/IV) and proximal or distal (relationship to splenic flexure). Effect of age was examined by stratifying patients into three groups (<50 years, 50-70 years, >70 years). Two time periods (1/87-12/96 and 1/97-12/00) were compared. Significance (P < 0.05) was determined by univariate and logistic regression analysis. Between January 1987 and December 2000, 1355 patients (M:F, 699:656; mean, 65.9 years) were entered into the tumor registry [998 whites, 357 African Americans (AA)]. The AA population had a significantly higher proportion of females (P = 0.0001) and patients <50 years (P = 0.01). The incidence of carcinoma in situ (CIS) was significantly higher in AA (P = 0.01). African Americans were more likely to present with late disease (P = 0.05), proximal cancers (P = 0.05), and well-differentiated tumors (P = 0.04). In the entire cohort, proximal lesions were significantly larger (P = 0.002), poorly differentiated (P = 0.002), and occurred more often in females (P = 0.03), patients >70 years (P = 0.04), and patients with family history of colon cancer compared to distal lesions. Proximal migration of tumors occurred in the latter part (1997-2000 compared to 1987-1996) of the study (P = 0.002). Patients <50 years had a higher incidence of late stage (P = 0.03) and poorly differentiated tumors (P = 0.009). The probability for a proximal tumor in an AA female >70 years was 61.9 per cent and in a white male >50 years was 35.1 per cent. Significant differences exist in the stage and location of tumors according to patient's age, race, and gender. These factors should be considered in implementing public screening strategies. Specifically, African-American patients were more likely to present with late-stage tumors, and more aggressive patient education and screening programs should be implemented. For all groups, a proximal migration of colorectal tumors was identified. This factor should eliminate use of sigmoidoscopy as a screening tool. Complete colonoscopy, instead, should be the procedure of choice to identify colonic neoplasia.