To evaluate the effects of sodium nitroprusside (SNP) and partial exsanguination (EXS) on systemic hemodynamics and cerebrospinal fluid dynamics, we monitored proximal and distal blood pressure (BP), cerebrospinal fluid pressure (CSFP), spinal cord perfusion pressure (SCPP), and compliance of the spinal canal (CSC) in 10 mongrel dogs during aortic cross-clamping of the descending thoracic aorta. CSC was measured by serial injections of 2 ml of saline solution into the cisterna cerebellomedullaris via a percutaneously placed catheter with simultaneous measurements of CSFP. Data were acquired at baseline (BL), during aortic cross-clamping with proximal hypertension (AXC), and after control of proximal hypertension with EXS and SNP. During the cross-clamp interval, mean proximal aortic pressure (PxBP) rose from 114 ± 6 to 150 ± 3 mm Hg ( P < 0.001), whereas mean blood pressure decreased to 88 ± 5 and 82 ± 4 mm Hg during the SNP and EXS intervals, respectively ( P < 0.05 vs BL). EXS and SNP were equally effective in controlling PxBP (82 ± 4 vs 88 ± 5 mm Hg, P > 0.05). Mean distal aortic pressure (DsBP) decreased from systemic values to 21.5 ± 1.9 mm Hg during AXC, and to 12.4 ± 1.0 and to 8 ± 0.8 mm Hg during EXS and SNP, respectively ( P < 0.05 AXC vs EXS and SNP). SNP lowered DsBP significantly more than EXS, 8 ± 0.8 vs 12.4 ± 1.0 mm Hg ( P < 0.05). CSFP decreased significantly during EXS compared with BL, AXC, and SNP (4.0 ± 1.32 vs 7.0 ± 0.87, 12.5 ± 1.43, and 18.8 ± 1.46 mm Hg, respectively, P < 0.05). As a result of the detrimental effects of SNP on DsBP and CSFP, SCPP was significantly lower during the infusion of SNP than during AXC and EXS (−10 ± 1 vs 10 ± 2 and 8 ± 1 mm Hg, respectively, P < 0.05). CSC increased significantly during EXS as compared with AXC (0.69 ± 0.04 to 1.15 ± 0.24 ml/mm Hg) whereas it decreased, although not significantly, during SNP compared with AXC and BL (0.55 ± 0.8 vs 1.15 ± 0.24 and 0.69 ± 0.04 ml/mm Hg, respectively, P > 0.05; AXC vs BL, P > 0.05). This study shows that the decrease in SCPP caused by SNP is caused by an increase in CSFP and a decrease in DsBP: however, the rise in CSFP does not appear to be caused solely by a decrease in CSC. The effects of reduction in CSFP observed with EXS are associated with an increase in CSC. Based on the results of this study, we conclude that the beneficial effects of EXS on SCPP may be mediated by an improvement in CSC.