Colonoscopic screening has been recommended for all persons who have had a colorectal adenoma or carcinoma. Such persons have been assumed to be at increased risk of having additional, asymptomatic colorectal neoplasms, the removal of which would reduce morbidity and mortality from colorectal cancer. In this prospective study, initial colonoscopy was performed on 544 asymptomatic subjects with past histories of colorectal index lesions ranging from small tubular adenomas to invasive cancers. In 402 subjects whose worst index lesion was an adenoma, the prevalence of neoplasms detected at colonoscopy, above the reach of the rigid sigmoidoscope, increased with age, male sex, black race, and the number and size of their index adenomas. In 142 subjects whose worst index lesion was invasive cancer, colonoscopy findings were marginally related to age and white race. A subgroup of 133 subjects whose worst index lesion was a single, small (less than 10 mm) tubular adenoma and who had no first-degree relatives with colorectal cancer had only a 3% prevalence of advanced colonic neoplasms (tubular adenomas greater than or equal to 10 mm in diameter; tubulovillous, villous, or severely dysplastic adenomas; or invasive cancers) found on colonoscopy--no greater than would be expected in the general population. Subgroups of the remaining 411 subjects, who had advanced or multiple index lesions, had prevalences of advanced neoplasms ranging from 8% to 18%. These findings indicate that for persons whose only risk factor is a single small tubular adenoma, current screening guidelines could be modified to recommend techniques less costly and less invasive than colonoscopy.