Obstetric endoscopy procedures are routinely performed at our institution to treat selected complications of monochorionic twin gestation. We perform these procedures under combined spinal epidural anesthesia plus maternal sedation. In the absence of general anesthesia, fetal immobilization is not achieved. We hypothesized that remifentanil would induce adequate maternal sedation and provide fetal immobilization, which is equal or superior to that induced by diazepam. Fifty-four second trimester pregnant women were included in this randomized, double-blind trial. After combined spinal epidural anesthesia, maternal sedation was initiated using either incremental doses of diazepam or a continuous infusion of remifentanil. Maternal sedation, hemodynamics, side effects, and fetal hemodynamics and immobilization were evaluated before, during, and for 60 min after surgery. Remifentanil produced adequate maternal sedation with mild but clinically irrelevant respiratory depression (respiratory rate 13 +/- 4 breaths/min and Pco(2) 38.6 +/- 4 mm Hg at 40 min of surgery), whereas diazepam resulted in a more pronounced maternal sedation but no respiratory depression (respiratory rate 18 +/- 3 breaths/min and Pco(2) 32.7 +/- 3 mm Hg at 40 min of surgery). Compared with diazepam, fetal immobilization with remifentanil occurred faster and was more pronounced, resulting in improved surgical conditions; the number of gross body and limb movements was 12 +/- 4 (diazepam) versus 2 +/- 1 (remifentanil) at 40 min of surgery. Because of this, the mean (range) duration of surgery was significantly shorter in the remifentanil-treated patients, 60 (54-71) min versus 80 (60-90) min in the diazepam group. We conclude that remifentanil produces improved fetal immobilization with good maternal sedation and only minimal effects on maternal respiration.