THE PROBLEM of the infant or child with rectal bleeding is familiar to all clinicians. The term "rectal bleeding" has through common usage come to mean the passage of blood from the rectum–whether it be bright red or dark, unrelated to, on the surface of, or mixed with the stool. Anal or gastrointestinal bleeding would be a more accurate term, but for the purposes of this paper rectal bleeding will be used in its widest connotation. This study was undertaken to attempt to evaluate the place of surgery in the investigation and treatment of this perplexing condition. The literature is surprisingly barren of any series of this kind. Several authors of textbooks on pediatric surgery do not even mention the possibility of a fruitless abdominal exploration. Others appear to hazard a guess. However, Nixon and O'Donnell state that probably one-third of patients with rectal bleeding never have the cause diagnosed. White and Dennison give 20% as a likely figure. Gross, in a small review of 65 patients presenting with rectal bleeding as a diagnostic problem, could find no cause for the bleeding in 18, i.e., 27%. Of the 18 patients who were operated upon, only 4 had negative laparotomies. Nevertheless, he states that in an astounding proportion of cases the cause for the rectal bleeding is never found, even after extensive investigation and laparotomy. Macbeth in a recent paper refers to the difficulties in gathering a series of cases with available methods of coding. Fortunately, at The Hospital for Sick Children, Toronto, case notes are coded in such a way that the histories of patients with rectal bleeding are readily available.