In order to find out how the need for analgesic supplementation during N2O-O2-relaxant anaesthesia is affected by chronic alcohol consumption, 82 patients with various known alcohol habits were anaesthetized for gastric or biliary surgery. Muscular relaxation was kept constant with the aid of a peripheral neurostimulator, and fentanyl was given in increments of 0.05-0.1 mg for nociceptive symptoms during the anaesthesia. For induction, alcoholics (annual consumption above 151 pure alcohol) needed significantly more thiopental/kg body weight than social drinkers (1--151 annually) or non-alcoholics (less than 11 annually), and excitative symptoms were more frequent in alcoholics. A positive correlation was found between fentanyl supplementation and alcohol consumption (r = 0.41), non-alcoholics requiring on average 3.8 microng/kg/h of fentanyl, as compared with 6.4 microng/kg/h in alcoholics. This correlation was clearer than that found previously under similar conditions by the authors between halothane supplementation and alcohol consumption (r = 0.20). In both studies the correlation was partly due to the higher incidence of gastric surgery among alcoholics, since gastric surgery itself requires more supplementation. An analysis of the different symptoms pointing to the need for fentanyl supplementation revealed that the simultaneous occurrence of several symptoms and the prevalence of motor responses were more common in alcoholics. In one patient halothane had to be used as well. In other patients no special difficulties were observed, and none of the patients reported dreams or recollections. The results suggest that during N2O-O2-relaxant anaesthesia the demand for fentanyl supplementation is increased by about 70% in chronic alcoholics with a mean annual consumption of 311 pure alcohol.