The demand for intermittent halothane supplementation during N2O‐O2‐relaxant anaesthesia was studied in 25 alcoholics (annual consumption over 15 1 pure alcohol) scheduled for biliary or gastric surgery. The controls were 45 nonalcoholics and 43 patients with an annual consumption of between 1 to 15 1. Thiopental (3 mg/kg/min) was given for induction. After intubation, halothane supplementation was given in 0.5% concentration for 10‐min periods. Standardized criteria for halothane supplementation were various motor and autonomic responses to painful stimuli. Muscular relaxation was kept fairly constant (roughly 90%), as assessed visually with the aid of a peripheral nerve stimulator. The total time for which halothane supplementation was given, expressed as a percentage of the total anaesthesia time, was used as an indication of the need for halothane supplementation. The need for thiopental for induction was not increased to a statistically significant extent in alcoholics, but signs of excitation did occur in 40% as compared with 11% in non‐alcoholics (P<0.01). The demand for halothane supplementation was higher in alcoholics (47±4.8%, s.e. mean) than in non‐alcoholics (33±2.3%). This difference, however, was partly due to the higher incidence of gastric surgery, which required more supplementation than biliary surgery. Analysis of the different criteria indicating the need for halothane supplementation revealed that an increase in blood pressure or heart rate was more common in non‐alcoholics, whereas motor irritability, sweating and lacrimation were more frequent in alcoholics. Management of the anaesthetic posed no special difficulties in the alcoholics with an estimated mean annual consumption of 32 ± 4 (s.e. mean) litres of absolute alcohol. Three patients (5% of the alcohol consumers) reported dreams or recollections, suggesting that this mode of halothane supplementation does not guarantee an adequate anaesthetic depth. The difficulties and biases associated with this type of analysis