Objective: Critical illness-related colonic ileus (CIRCI) is characterized by the non-passage of stools in critically ill patients as a result of the absence of prokinetic movements of the colon, while the upper gastrointestinal tract functions properly and mechanical ileus is absent. We investigated whether neostigmine resulted in defecation in patients with CIRCI. Design: Double-blinded, placebo-controlled prospective study. Setting: Eighteen-bed intensive care unit. Patients: Thirty ventilated patients with multiple organ failure with CIRCI for > 3 days. Intervention: Continuous intravenous administration of neostigmine 0.4-0.8 mg/h over 24 h, or placebo. Measurements and results: Time to first defecation and adverse reactions were recorded. Thirty patients were randomized, 24 could be evaluated. The mean prestudy time was 5 days, mean APACHE II score on admission was 23.2, and mean MOF score on the day of the study was 6.4. Of the 13 patients receiving neostigmine, 11 passed stools, whereas none of the placebo-treated patients passed stools (P < 0.001). After 24 h, the non-responders received in a cross-over fashion neostigmine or placebo respectively. Eight out of the 11 neostigmine patients now passed stools (mean 11.4 h), and none of the placebo patients. Overall, in none of the patients did passage of stools occur during placebo infusion, whereas 19 of the 24 neostigmine-treated patients had defecation (79%). No acute serious adverse effects occurred, but three patients had ischemic colonic complications 7-10 days after treatment. Conclusion: Continuous infusion of 0.4-0.8 mg/h of neostigmine promotes defecation in ICU patients with a colonic ileus without important adverse reactions.