OBJECTIVE Early defecation after reconstructive anal surgery may influence the outcome negatively. Different methods are used to avoid bowel movements in the early postoperative period. We questioned whether stool behaviour is influenced by total parenteral nutrition as opposed to enteral nutrition with resorbable sip feeds. Furthermore, satisfaction of patients with each nutrition regime, cost differences and influence of the postoperative outcome were evaluated. METHODS Between January and October 2004, 32 patients were evaluated in a prospective randomized, surgeon-blinded trial. The parenteral group (PG, n = 16) received 1250 ml Nutriflex lipid basal (B. Braun Comp., Melsungen, Germany) intravenously. The enteral group (EG, n = 16) was offered a total amount of three cups of Clinutren fruit (Nestle Nutrition GmbH, Frankfurt, Germany), two boxes of OPD (oligopeptide diet) Elemental 028 extra (SHS, Liverpool, UK) and two cups of OPD Peptamen (Nestle Nutrition GmbH). Both groups received 1000 ml of isotonic cristalloid solution and were allowed to drink up to 1000 ml of tea or water per 24 h. RESULTS The satisfaction of patients was significantly higher in the EG. In regard to stool behaviour both groups showed no differences, in the number of bowel movements or in the time to the first postoperative defecation. Enteral feeding resulted in a minimum saving of 220 euros. Postoperative results in the case of plastic fistula closure did not differ between both groups. CONCLUSIONS Sufficient bowel confinement during the early postoperative period after anal reconstructive surgery may be achieved by using resorbable sip feeds rather than parenteral nutrition.