PURPOSE: The aim of this study was to assess any differences between the inclusion or omission of medical bowel confinement relative to postoperative morbidity and patient tolerance after anorectal reconstructive surgery. METHODS: Between January 1995 and February 1997 a prospective randomized trial was conducted for patients without stomas who underwent anorectal reconstructive surgery. All patients were randomly assigned either to medical bowel confinement (a clear liquid diet with loperamide 4 mg by mouth three times per day and codeine phosphate 30 mg by mouth four times per day until the third postoperative day) or to a regular diet, beginning the day of surgery. AU patients in both groups underwent the identical preoperative oral mechanical preparation, preoperative oral and parenteral antibiotics, and postoperative antibiotics. Wound closure and wound care were identical in both groups. RESULTS: Fifty-four patients (46 females) were prospectively, randomly assigned to medical bowel confinement (n = 27; 50 percent) or a regular diet (n = 27; 50 percent); the mean ages were 51.0 (range, 28-80) and 47.2 (range, 23-87) years, respectively. Indications for surgery were fecal incontinence in 32 patients, complicated fistulas in 17 patients, anal stenosis in 4 patients, a Whitehead deformity in 1 patient, and a chronic unhealed fissure in 1 patient. Fifty-four patients underwent 55 procedures: 32 patients underwent sphincteroplasty, 18 patients underwent transanal advancement flaps, and 5 patients underwent anoplasties. There were no differences between the two groups in the incidence of either septic or urologic complications. Nausea and vomiting were recorded in seven (26 percent) medical bowel confinement and three (11 percent) regular-diet patients. The first postoperative bowel movement occurred at a mean of 3.9 days in the medical bowel confinement group and 2.8 days in the regular diet group (P < 0.05). Fecal impaction occurred in seven (26 percent) of the patients in the medical bowel confinement group and two (7 percent) of the patients in the regular diet group. Hospital charges analysis showed a mean cost of hospitalization of $12,586.00 (range, $3,436.00-$20,375.00) for the medical bowel confinement group and $10,685.00 (range, $3,954.00-$18,574.00) in the regular diet group, representing a mean difference of $1,901.00 (P = 0.06). Mean follow-up was 13 months for both groups (range, 1-24 months in the regular diet group and 2-25 months in the medical bowel confinement group). No statistical difference was shown in the functional outcome of sphincteroplasties between the medical bowel confinement group and the regular diet group. CONCLUSIONS: The outcome of reconstructive anorectal surgery was not adversely affected by the omission of medical bowel confinement. Moreover, cost savings can be achieved by the omission of routine bowel confinement.