Summary: A multicentre randomized trial of 10 days levamisole 2 mg/kg given in three doses daily, starting 7 days before surgery and proceeding as early after surgery as possible was performed. All patients were preoperative evaluated. Known prognostic factors were registered. The clinical status and CEA serum concentration were registered every 3 months. The survival curves as well as clinical relapse and laboratory relapse were determined separately for the groups of perioperative levamisol treated and untreated patients. Statistical significance was evaluated by Cox's regression analysis with consideration given to all known prognostically significant factors. There was no difference in the survival rate between treated and untreated groups.