Recent clinical reports have suggested that drainage of cerebrospinal fluid lowers the incidence of perioperative paraplegia in patients with thoracoabdominal aneurysms. Unfortunately, the precise mechanisms for both the neurologic deficits and the beneficial effects of cerebrospinal fluid drainage remain unclear. To better understand the relationship between cerebrospinal fluid pressure, central venous pressure, and the compliance of the cerebrospinal fluid compartment, we studied 12 anesthetized dogs subjected to thoracic aortic occlusion. Pericardia were opened in six (group I), and left intact in six (group II). Systemic hemodynamics and cerebrospinal fluid pressure (mm Hg) were measured before and after thoracic aortic occlusion. In group II, intravenous volume loading (15 ml/kg) was superimposed on aortic occlusion. Compliance of the cerebrospinal fluid space (ml/mm Hg) was measured at each interval by use of sequential injection and withdrawal of small aliquots of fluid. Results are expressed as mean +/- SE; *p less than 0.05. Thoracic aortic occlusion resulted in predictable changes in mean arterial pressure (group I 95.8 +/- 7.1 to 123.3 +/- 7.1*, group II 82.5 +/- 6.9 to 98.3 +/- 9.5*) and central venous pressure (1.9 +/- 0.7 to 3.8 +/- 0.6*, 3.0 +/- 0.8 to 4.0 +/- 0.9*). Although cerebrospinal fluid pressure was increased by thoracic aortic occlusion in both groups (8.0 +/- 1.2 to 12.6 +/- 1.9*; 5.8 +/- 0.9 to 8.5 +/- 1.1*), compliance of the dural space was was not changed (0.61 +/- 0.19 to 0.60 +/- 0.18; 0.54 +/- 0.14 to 0.62 +/- 0.09).(ABSTRACT TRUNCATED AT 250 WORDS)