In dogs anesthetized with pentobarbital-chloralose, cardiac output and blood flows of four regional vascular beds (superior mesenteric, left renal, left circumflex coronary and left femoral) were continuously monitered with electromagnetic flowmeters. Arterial blood pressure and heart rate were also measured. Hypotension was induced with intravenous infusions of sodium nitroprusside and trimethaphan for 5-16 min to produce comparable reductions of mean arterial pressure (32 mm Hg or 26 per cent with nitroprusside and 37 mm Hg or 31 per cent with trimethaphan). Cardiac output also decreased, but to a lesser extent (11.5 per cent with nitroprusside and 12.5 per cent with trimethaphan). Thus, total peripheral resistance was consistently decreased. Nitroprusside caused slight tachycardia, while trimethaphan produced bradycardia. Both drugs decreased mesenteric blood flow and increased mesenteric vascular resistance. Renal blood flow was maintained or increased with nitroprusside; thus, renal vascular resistance decreased; with trimethaphan, renal blood flow decreased and renal vascular resistance did not change. Both nitroprusside and trimethaphan reduced coronary blood flow; the reduction was more pronounced with the latter. Nitroprusside affected femoral blood flow minimally, with a slight reduction of femoral vascular resistance. In contrast, trimethaphan increased femoral blood flow and markedly decreased femoral vascular resistance. Redistribution of cardiac output favoring the dilated skin and muscle vascular beds appears to be an important undesirable effect of trimethaphan.