BACKGROUND: Nausea and vomiting during regional anesthesia for a cesarean section are common and unpleasant complications. This study was designed to assess the efficacy of prophylactic low dose granisetron on nausea and vomiting of patients undergoing a cesarean section with epidural anesthesia in a prospective, randomized, and double blind manner. METHODS: Sixty obstetric patients not in labor received epidural anesthesia with 20 mL of 2% lidocaine plus epinephrine (1 : 100,000) and 4 mL of diluted morphine 2 mg was administrated for postoperative pain control. Maternal hypotension was prevented aggressively by fluid administration and an ephedrine injection. The patients were assigned randomly to one of four groups to receive a placebo (normal saline) or granisetron in a dose of 10micro gram/kg, 20micro gram/kg, or 30micro gram/kg i.v. before peritoneal closure. Emetic episodes and emetic scores rating from 0 to 4 were observed during the 24 hours after surgery. RESULTS: The emetic score was statistically lower in the granisetron 30micro gram/kg group than the other groups, and was lower in the granisetron 20micro gram/kg group than the placebo and granisetron 10micro gram/kg group. Adverse events were headache, dizziness, sedation and pruritus, which were not serious. CONCLUSIONS: Prophylactic use of low dose granisetron (20-30micro gram/kg) is effective in postoperative nausea and vomiting in a cesarean section with epidural anesthesia using lidocaine and morphine when hypotension as one of the contributing factors was minimized.