In a 22-year period, 120 patients were admitted for treatment of a high imperforate anus. Sixty-one patients were available for long-term assessment following a pull-through procedure. These patients were evaluated by written questionnaire and personal interview. Both a qualitative and quantitative fecal continence score were established to provide a system for long-term follow-up that was both reproducible and easy to use. These results were grouped according to the length of time following repair and type of repair performed. Three follow-up time periods were used: 2.5 to 9 years; 10 to 16 years; and 17 to 24 years. During this study, three different procedures were used: sacroperineal (SP), sacroabdominoperineal (SAP), and abdominoperineal (AP). Grouping all three procedures together, the percentage of patients with "good" fecal continence increased in time from 33% to 58% to 64%. Patients undergoing an SP repair uniformly had good results by 10 or more years following repair. In contrast, all patients undergoing an SAP or AP repair had only fair quantitative results 10 to 16 years following repair. Females generally had much better results than males but females also tended to have a lower rectal pouch than males. A comparison of this survey with the results of Swenson, Kiesewetter, Stephens, and Rehbein supports the following conclusions: the anatomic level of rectal pouch at birth predicts fecal continence better than the sex of the patient or type of procedure used; patients in whom the full thickness, terminal rectal wall is used in performing the pull-through procedure have the best fecal continence results.(ABSTRACT TRUNCATED AT 250 WORDS)