OBJECTIVE The aim of this study was to compare morphine consumption with patient-controlled analgesia (PCA) between spinal anesthesia (SA) (bupivacaine, morphine and fentanyl) and general anesthesia (GA) (sufentanil) after an abdominal hysterectomy. METHODS Forty women were randomly assigned to receive SA with bupivacaine 15 mg, 0.15 mg of intrathecal morphine and 15 microg of fentanyl or GA with sufentanil, both combined with PCA. The primary outcome was morphine consumption with the PCA device. The secondary outcomes were post-operative pain at rest and under stress on a visual analog scale, nausea, pruritus and respiratory depression on a standardized scale. Outcome measures were recorded at 6, 12, 18, 24 and 48 h post-anesthesia. The duration of post-anesthesia care unit (PACU) and hospital stay were recorded. RESULTS Patients in the SA group consumed at least two times less morphine at each time interval than the GA group: at 48 h, they used 19 +/- 17 vs. 81 +/- 31 mg (P<0.0001). Post-operative pain at rest was lower in the SA group until the 18th hour and under stress until the 48th. There was more sedation in the GA group until the 18th hour. Little difference was observed in the incidence of pruritus. Nausea was more intense at the 6th hour in the GA group. There was no difference in the respiratory rate. The duration of PACU stay was shorter for the SA group (52 +/- 9 vs. 73 +/- 11 min, P<0.0001) as was the duration of hospital stay (2.2 +/- 0.4 vs. 3.3 +/- 0.7 days, P=0.01). CONCLUSIONS It is concluded that intrathecal morphine 0.15 mg with 15 microg of fentanyl decreases post-operative pain and morphine consumption by PCA without increasing adverse reactions for women undergoing an abdominal hysterectomy.