Fifty-seven patients, all over the age of 64, with femoral neck fracture were randomized to receive epidural or halothane anesthesia to see if the anesthetic technique influenced the incidence of postoperative confusion. All patients were lucid on admission. Using the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-III) as criteria for confusion, we found that 44% of the patients developed confusion that correlated closely to a history of mental depression (P less than 0.01) and to the use of drugs with anticholinergic effect (P less than 0.005). There was no difference in the incidence of confusion between the two anesthetic groups. In patients given halothane, however, early postoperative hypoxemia was associated with confusion (P less than 0.05). Patients with confusion had significantly more postoperative complications and almost four times longer hospitalization times. It is concluded that anticholinergic medication and a history of mental depression are predominant risk factors for development of postoperative confusion and in this respect are more important than the anesthetic technique.