Cerebrospinal fluid (CSF) drainage has been reported to protect the spinal cord during surgical procedures requiring thoracic aortic cross-clamping. In 1986, CSF pressure monitoring and drainage was begun in an attempt to reduce the incidence of paraplegia associated with surgical repair of descending thoracic and thoracoabdominal aortic aneurysms (TAAA). These Group II patients (n = 50) were retrospectively compared to Group I patients (n = 49) who had undergone similar surgical procedures in the previous 3 years before CSF monitoring was introduced into this practice. Group II patients had intrathecal catheters placed for monitoring of CSF pressure and drainage of CSF to maintain the pressure < or = 15 mmHg. Seven patients (four in Group I, three in Group II) died before recovering from the anesthetic. Of the 47 patients in Group II who survived, none had clinically apparent complications such as an epidural hematoma or meningitis from the intrathecal catheter. The mean aortic cross-clamp time was 58.6 +/- 30.5 minutes (mean +/- SD) in Group I versus 65 +/- 42.6 minutes in Group II. Twenty-three patients in Group I and 16 patients in Group II had a shunt to the distal aorta. To maintain a CSF pressure of < or = 15 mmHg in Group II, an average of 46.9 +/- 6.9 mL of CSF was withdrawn. Of the 45 survivors in Group I, 4 developed a spinal cord deficit; the number of patients with spinal cord deficit in Group II was 4 out of the 47 who survived.(ABSTRACT TRUNCATED AT 250 WORDS)