Barbiturates induce reduction of myocardial contractility and metabolism, whereas ketamine exerts a sympathomimetic effect that can mask its direct depressant effect on contractility. However, it is unclear whether barbiturates, which interfere with the cytochrome P-450 pathway, or ketamine, which inhibits nitric oxide synthesis, also alter the responsiveness of the coronary vessels to vasodilator stimuli. We hypothesized that the parameters of coronary reactive hyperemia (CRH), which reflect both the degree of myocardial metabolism and vascular reactivity, could be modified by the type of anesthesia used. In two groups of goats, anesthesia was induced either using ketamine plus nitrous oxide or pentobarbital alone. To record coronary flow an electromagnetic flow-probe was placed around the left circumflex coronary artery. In the ketamine group (n = 14) and in the pentobarbital group (n = 16) CRH was studied using the indices of myocardial metabolism and vascular dilator responsiveness. In the pentobarbital group all of the indices of myocardial metabolism were lower than in the ketamine group (i.e. the excess to debt flow ratio was 2.3+/-0.8 vs. 4.6+/-2.4; p< 0.001). Yet, some indices of vascular responsiveness (time derivative of coronary flow and the peak to basal flow ratio) were not different in the two groups. Moreover, the duration of the reactive hyperemia was shorter in the ketamine than in the pentobarbital group (118+/-47 vs. 153+/-45 s, p<0.05). It is suggested that pentobarbital decreases the indices of CRH related to metabolic activity, whereas ketamine reduces the duration of the hyperemic response, which suggests an impairment of endothelial function.