Most obstetric anesthesiologist now consider epidural anesthesia to be the preferred anesthetic technique for patients with preeclampsia. In the severe preeclampsia effective ciirculatory volume is usually low, so appropriate preanesthetic hydration is essential feature of safe management. We performed CVP monitoring before epidural anesthesia in nine severe preeclamptic patients undergoing cesarean section. Mean CVP of them was 2+/-1 mm Hg which was not significantly different from normal term parturient. In four patients whose CVP was less than 2 mmHg, preanesthetic hydration with Hartman's solution 500~700 rrd was performed under the monitoring of CVP until it became to rise 2~3 mmHg. Five patients out of nine whose systolic blood pressure dropped below 100 mmHg during anesthesia were IV injected ephedrine 4 mg immmediately and three of them needed repeat injection of ephedrine 4 mg. With the monitoring of CVP, the operations of all the nine severe preeclamptic patients were completed successfully without any problem during and after anesthesia. lt is important to reaUze the limitation of central venous pressure monitoring : but if used with caution in proper circumstance it can provide clinically useful information with minimal risk to the patient.