The present study was designed to determine whether the Cordis lumbar catheter is more effective than spinal needles or epidural catheters for intraoperative removal of cerebrospinal fluid (CSF). We also sought to determine whether placing two spinal needles, two epidural catheters, or two Cordis CSF catheters per patient provides more effective removal of CSF than placing one such CSF drain. Spinal needles were associated with a higher incidence of fresh blood in the CSF when CSF removal was desired, a lower incidence of adequate volume of CSF drainage, and a higher incidence of the need to reposition needles to continue CSF removal. Epidural catheters were associated with a lower incidence of adequate rate of CSF drainage. The slow rate of CSF drainage with epidural catheters was clinically important because it impaired operative conditions, delayed surgical progress, or both. Cordis CSF catheters took longer to place than spinal needles or epidural catheters. This time difference (about 3 minutes for one CSF catheter and 6 minutes for two) was statistically significant, but probably is not clinically important. The results were similar whether one or two CSF drains were used per patient. It is concluded that Cordis CSF catheters are more effective than spinal needles or epidural catheters for intraoperative removal of CSF. Using two spinal needles, two epidural catheters, or two Cordis CSF catheters per patient did not provide better CSF removal than using one such drain.