BACKGROUND This study was designed to examine whether the coupling between oxygen consumption (VO2) and cardiac output (CO) is maintained during xenon anaesthesia. METHODS We studied the relationship between VO2 (indirect calorimetry) and CO (ultrasound flowmetry) by adding xenon to isoflurane anaesthesia in five chronically instrumented dogs. Different mixtures of xenon (70% and 50%) and isoflurane (0-1.4%) were compared with isoflurane alone (1.4% and 2.8%). In addition, the autonomic nervous system was blocked (using hexamethonium) to study its influence on VO2 and CO during xenon anaesthesia. RESULTS Mean (SEM) VO2 increased from 3.4 (0.1) ml kg(-1) min(-1) during 1.4% isoflurane to 3.7 (0.2) and 4.0 (0.1) ml kg(-1) min(-1) after addition of 70% and 50% xenon, respectively (P<0.05), whereas CO and arterial pressure remained essentially unchanged. In contrast, 2.8% isoflurane reduced both, VO2 [from 3.4 (0.1) to 3.1 (0.1) ml kg(-1) min(-1)] and CO [from 96 (5) to 70 (3) ml kg(-1) min(-1)] (P<0.05). VO2 and CO correlated closely during isoflurane anaesthesia alone and also in the presence of xenon (r2=0.94 and 0.97, respectively), but the regression lines relating CO to VO2 differed significantly between conditions, with the line in the presence of xenon showing a 0.3-0.6 ml kg(-1) min(-1) greater VO2 for any given CO. Following ganglionic blockade, 50% and 70% xenon elicited a similar increase in VO2, while CO and blood pressure were unchanged. CONCLUSIONS Metabolic regulation of blood flow is maintained during xenon anaesthesia, but cardiovascular stability is accompanied by increased VO2. The increase in VO2 is independent of the autonomic nervous system and is probably caused by direct stimulation of the cellular metabolic rate.