This study compares intrathecal (IT) fentanyl with IV ondansetron for preventing intraoperative nausea and vomiting during cesarean deliveries performed with spinal anesthesia. Thirty healthy parturients presenting for elective cesarean delivery with standardized bupivacaine spinal anesthesia were randomized to receive 20 microg IT fentanyl (Group F) or 4 mg IV ondansetron (Group O) by using double-blinded methodology. At eight specific intervals during the surgery, a blinded observer questioned the patient about nausea (1 = nausea, 0 = no nausea), observed for the presence of retching or vomiting (1 = vomiting or retching, 0 = no vomiting or retching), and recorded a verbal pain score (0-10, 0 = no pain, 10 = worst pain imaginable). Cumulative nausea, vomiting, and pain scores were calculated as the sum of the eight measurements. Intraoperative nausea was decreased in the IT fentanyl group compared with the IV ondansetron group: the median (interquartile range) difference in nausea scores was 1 (1, 2), P = 0.03. The incidence of vomiting and treatment for vomiting was not different (P = 0.7). The IT fentanyl group had a lower cumulative perioperative pain score than the IV ondansetron group; the median difference in the cumulative pain score was 12 (8, 16) (P = 0.0007). The IT fentanyl group required less supplementary intraoperative analgesia. The median difference in the cumulative fentanyl dose was 100 (75, 100) microg fentanyl, (P = 0.0002).