Background: Cardiopulmonary adverse effects are commonly observed in patients undergoing colonoscopy with sedation. This study determines the incidence of such events when sedation is given only when required. Methods: In 2500 consecutive patients, colonoscopies were started without premedication but sedation was offered if significant discomfort occurred. Parameters of blood pressure, heart rate, oxygen saturation and patient appearance were continuously recorded by a nurse assistant. Additional adverse effects occurring during or after the procedure were noted in the patients' protocol. Risk factors for the development of complications were evaluated with the use of a logistic regression model and the odds ratio. Results: Ninety-five percent of all patients required neither sedation nor analgesia. Adverse effects occurred in 59 patients (2.4%). Twenty-six of these patients (1.0%) had short-lasting episodes of oxygen desaturation and 22 patients (0.9%) experienced vasovagal reactions. Mechanical complications occurred in a total of 8 patients (0.3%) and consisted of 2 perforations and 6 episodes of hemorrhage. In the logistic regression model, impaired physical status was the single most important risk factor for the development of cardiopulmonary complications (odds ratio 4.7; 95% confidence interval [2.0, 11.4]). Conclusions: In experienced hands, patients undergoing colonoscopy rarely require sedation. If selective sedation is used, cardiopulmonary adverse effects occur in approximately 2% of all patients, most of whom require no medical intervention. (Gastrointest Endosc 1999;49:560-5.)