To determine the clinical significance of finding occult blood in a stool sample obtained during digital rectal examination of patients with no gastrointestinal complaints, we reviewed the records of patients who had colonoscopy for this indication. Of the 185 patients (average age, 59.4 years) who met study criteria, 48 were inpatients when the Hemoccult-positive rectal examination was recorded. Neoplastic lesions were found in 28% of the patients (51/185). Thirteen of the lesions were adenocarcinomas and 38 were adenomatous polyps. Seventy-five patients (41%) had normal findings on colonoscopy. Nonneoplastic lesions were found in the remaining 32%. Comparison of patients with neoplastic lesions and patients with nonneoplastic lesions showed no significant difference with respect to the presence of anemia (43% vs 48%), inpatient status at the time of positive rectal examination (23% vs 29%), or sex. Although the mean age was similar in patients with and without neoplastic lesions (60.5 and 56.8 years, respectively), no neoplastic lesions were found in patients less than 45 years old. We conclude that testing stool obtained during rectal examination for occult blood does not increase the rate of false-positive results, and positive test should prompt a structural evaluation of the colon to detect neoplasia in patients with an age-related risk for colon carcinoma.