Background: The aim of this systematic review was to compare the safety and efficacy of dynamic graciloplasty with colostomy for the treatment of faecal incontinence. Methods: Two search strategies were devised to retrieve literature from the Medline, Current Contents, Embase and Cochrane Library databases up until November 1999. Inclusion of papers depended on a predetermined protocol, independent assessments by two reviewers and a final consensus decision. English language papers were selected. Acceptable study designs included randomized controlled trials, controlled clinical trials and case series. Forty papers met the inclusion criteria. They were tabulated and critically appraised in terms of methodology and design, outcomes, and the possible influence of bias, confounding and chance. Results: No high-level evidence was available and there were no comparative studies. Mortality rates were around 2 per cent for both graciloplasty and colostomy. Morbidity rates reported for graciloplasty appear to be higher than those for colostomy. Dynamic graciloplasty was clearly effective at restoring continence in between 42 and 85 per cent of patients, whereas colostomy is, by its design, incapable of restoring continence. However, dynamic graciloplasty is associated with a significant risk of reoperation. Conclusion: While dynamic graciloplasty appears to be associated with a higher rate of complications than colostomy, it is clearly a superior intervention for restoring continence in some patients. It is recommended that a comparative, but non-randomized, study be undertaken to evaluate the safety of dynamic graciloplasty in comparison to colostomy, and that the procedure should be performed only in centres where it is carried out routinely.