Sodium nitroprusside (SNP) has been used to control the proximal hypertension associated with thoracic aortic cross-clamping (TACC) during thoracic aortic surgery. It worsens neurologic outcome, presumably by further decreasing distal arterial pressure and increasing cerebrospinal fluid (CSF) pressure, thereby worsening the spinal cord perfusion pressure (SCPP). Trimethaphan does not increase CSF pressure. Therefore, the present study investigates the effect of trimethaphan versus SNP to control proximal hypertension during TACC on neurologic outcome. Two groups, each with eight mongrel dogs, were studied. All animals underwent descending TACC for 45 min. The mean proximal aortic blood pressure was maintained at 95-100 mm Hg by the use of SNP or trimethaphan. Distal aortic pressure was allowed to vary. The dogs were neurologically evaluated 24 and 48 h later by a blinded observer. During cross-clamping, there was no difference in mean proximal aortic pressure between groups. After 10 min of cross-clamping, distal aortic pressure was higher (P < 0.01), CSF pressure was lower (P < 0.01), and SCPP was higher (P < 0.005) in the trimethaphan group as compared with the SNP group (group effect). Neurologic outcome as assessed by Tarlov's score was better at 24 and 48 h in the trimethaphan group (P < 0.05). Histopathologic injury trended with worsened neurologic outcome. We conclude that 1) trimethaphan produced higher SCPP than SNP, and 2) neurologic outcome was better in the trimethaphan group.