BACKGROUND & AIMS: Dynamic muscle plasty has been advocated as therapy for refractory fecal incontinence and for anorectal reconstruction to avoid colostomy after abdominoperineal resection. This study evaluates the results of a multicenter experience with dynamic muscle plasty in the treatment of fecal incontinence and total anal reconstruction. METHODS: One hundred thirty-nine patients were enrolled at 12 centers between June 1992 and November 1994 and followed up through June 1996. Intramuscular leads and neurostimulators were implanted to stimulate transposed gracilis or gluteus muscle. Success was defined as 70% reduction in solid stool incontinence for patients with baseline incontinence and zero incontinence to solid stool for patients with baseline stomas and for patients undergoing total anal reconstruction. RESULTS: Overall, 85 of 128 graciloplasty patients (66%) achieved and maintained a successful outcome over the follow-up period. By etiology, these proportions were 71%, 50%, and 66% for patients with acquired fecal incontinence, congenital incontinence, and total anal reconstruction, respectively. One third of graciloplasty patients experienced a major wound complication, with therapy failing in 41%. Experienced centers had better outcomes and lower complication rates than inexperienced centers. Of the 11 gluteoplasty patients, 5 (45%) achieved and maintained a successful outcome. CONCLUSIONS: Dynamic graciloplasty may be an effective procedure for patients with refractory, end-stage fecal incontinence as well as for patients who require anorectal excision for low-lying malignancy. However, the procedure has significant morbidity that can lead to functional failure. Outcome after dynamic graciloplasty appears to correlate with surgical experience. In contrast to graciloplasty, the use of dynamic gluteoplasty should be limited to investigational purposes.