Ketamine is a racemic mixture containing equal amounts of optical isomers that have almost identical pharmacokinetic properties but different pharmacodynamic effects. The S-(+)-isomer of ketamine has about twice the anaesthetic and analgesic potency of the racemic ketamine preparation and is judged to induce less psychic emergence reactions and to be followed by a more rapid recovery of vigilance. The present study was designed to assess whether the S-(+)-isomer of ketamine is superior to the racemic mixture in cardiovascular characteristics, emergence reactions and cognitive functions, and whether side effects may be reduced or prevented by administration of midazolam prior to injection of S-(+)-ketamine. METHODS. Following ethics committee approval and informed consent, 30 volunteers were randomly allocated in this double-blind study to three groups of 10 each. Group 1 received 2 mg/kg bw racemic ketamine, group 2, 1 mg/kg bw S-(+)-ketamine and group 3, 1 mg/kg bw S-(+)-ketamine after premedication with 0.1 mg/kg midazolam i.v. Cardiovascular changes, state of vigilance, cognitive performance, subjective mood and acceptance of anaesthesia were assessed by means of haemodynamic routine monitoring, electroencephalography (EEG), psychometric tests and interview. RESULTS. The increases in mean arterial pressure and heart rate following the injection of racemic ketamine and S-(+)-ketamine were identical and the differences from baseline values significant after both. Premedication with midazolam ensured stable haemodynamics after injection of S-(+)-ketamine. EEG analysis displayed the characteristic changes well known from ketamine anaesthesia for both racemic and S-(+)-ketamine. The vigilosomnoscript showed an identical profile of vigilance up to 30 min after injection of both drugs. The vigilance status after 125 min was less impaired by S-(+)-ketamine than by racemic ketamine. Psychological assessment showed a prompter recovery of visual attentiveness and sensorimotor performance in the S-(+)-ketamine group. Subjective mood was judged by the volunteers to be significantly better after S-(+)-ketamine, and volunteers found S-(+)-ketamine to be more acceptable than racemic ketamine. The frequency of dreams was the same after both drugs. No unpleasant dreams were reported after S-(+)-ketamine, but one of the volunteers who received racemic ketamine had uncomfortable dreams. Midazolam prevented any unpleasant emergence sequelae. On the other hand, the cognitive performance could not be restored to the baseline values until at least 240 min after injection of S-(+)-ketamine, because of the sedative effects of midazolam. DISCUSSION. These results suggest that S-(+)-ketamine offers the advantages of faster recovery of cognitive performance, greater acceptance by the volunteers and identical depth of anaesthesia after injection of half the dose compared with racemic ketamine. The clinical use of S-(+)-ketamine therefore seems to be justified. Premedication with benzodiazepines, e.g. midazolam, is essential. The dose to be administered, however, should be carefully selected in order not to abolish the positive effect of S-(+)-ketamine on vigilance by the sedative effects of the benzodiazepine.