Although worsening cerebral function in the elderly is often said to be due to anaesthesia, it still remains to be proved that anesthesia acts on an elderly patient's psychological make-up. This study aimed to compare the psychological effects of general or spinal anaesthesia on 35 patients more than 70-yr old under repair of a subcapital femoral fracture, after having excluded 63 patients. They were randomly divided into two groups, the first receiving a spinal anaesthetic with 1.5 mg.kg-1 prilocaine (RA; n = 19), and the others a general anaesthetic with 5 mg.kg-1 thiopentone, dextromoramide, nitrous oxide and enflurane (GA; n = 16). None of the patients wer given a premedication. They underwent a battery of six psychological tests before and 36 h after surgery, all carried out by the same physician. Six patients in GA group and three in RA group developed a transient fall in mean blood pressure just after anaesthetic induction (not significant). There were no significant differences between the two groups in age, sex distribution, ASA class and performance in the tests, both before and after surgery. Individual psychological scores did not differ either. In the RA group, the postoperative tests were even better carried out than before surgery (p less than 0.05). It could therefore be concluded that anaesthesia alone has little impact on the psychological status of elderly patients.