Continuous epidural anesthesia (CEA) is generally accepted as a routine method of regional anesthesia while there has been only limited application of continuous spinal anesthesia (CSA), due mainly to a lack of adequate spinal catheters. With the introduction of a new, ultra-thin spinal catheter (32 G) inserted via a thin puncture needle, some of the complications reported after CSA can be eliminated. We studied CSA versus CEA in lower-extremity operations. METHODS. We evaluated 33 patients in a prospective, randomized study. All were comparable with respect to age, anesthetic risk (ASA II-III), and pre-existing diseases. The only exclusion criterium was the presence of a coagulation disturbance. The CSA group consisted of 17 patients (mean age 75.5 +/- 0.1 year); 26 G puncture needle and 32 G catheter were used. The CEA group consisted of 16 patients (mean age 73.8 +/- 11.0 years); an 18 G puncture needle and 22 G epidural catheter with a stylet were inserted with the loss-of-resistance technique. Both catheters were placed with the patient in a sitting position and left in place for 24 h in order to administer local anesthetics (LA) for postoperative analgesia as required. Hemodynamic parameters-mean arterial pressure (MAP) and heart rate (HR)-were compared in each group at 5-min intervals for 30 min after administration of local anesthetic and at 10-min intervals during the operation. Additionally, the ECG, pulse oximetry, respiratory rate, diuresis, and blood gases were monitored. After placement of the catheter, patients in the CSA group received 1.9 ml (+/- 0.2) bupivacaine HCl 0.5%. Patients in the CEA group received 12.6 ml (+/- 2.5) bupivacaine HCl 0.5%. For statistical evaluation of the data we used mean values, standard deviation (+/-), the Kruscal-Wallis procedure, and Student's t-test for unpaired data. P less than 0.05 was considered significant. RESULTS. The mAPs in the CSA group generally remained lower than those of the CEA group. However, over the course of the operation as well as after repeated injections, the difference between the two groups decreased. Only at 5 min after administration of the initial dose was a statistically significant difference in blood pressures between the two groups observed. A clinically relevant, rapid decrease in blood pressure due to relatively high doses of LA was seen in 1 case in each group. The first reinjection of LA after the initial dose was after 1.9 h in the CSA group (bupivacaine HCl 0.5% 1 +/- 0.3 ml) and after 1.8 h in the CEA group (bupivacaine HCl 0.5% 4.5 +/- 1 ml). The total dose of bupivacaine in the CSA group was 0.18 ml/kg per hour versus 0.8 ml/kg in the CEA group. No post-dural puncture headache was observed in the CSA group. DISCUSSION. The catheter designed for CSA is easy to use, although because of its small diameter a certain manual dexterity is required. In addition, CSA resulted in a more rapid onset of action and more pronounced sensorimotor blockade than did CEA. Hemodynamic alterations and side effects were comparably low in both groups.