A double-blind, parallel group, randomized trial was carried out to compare the hypnotic efficacy, tolerance and psychometric effects of two doses of loprazolam and placebo in 89 elderly hospital in-patients. All had disturbed sleep patterns, which were not the result of serious psychiatric or physical illness and, after a 1 or 2 day single-blind placebo baseline period, they received either 0.5 mg or 1.0 mg loprazolam or placebo administered double-blind for 5 days. The sleep of patients in all three treatment groups, as measured by patient self-rating scales, substantially improved during the study, with loprazolam being superior to placebo for measures of sleep latency, satisfaction with sleep, number of nocturnal awakenings and feeling refreshed on waking. Differences in hypnotic efficacy between the 0.5 mg and 1.0 mg loprazolam groups were not significant. The incidence of side-effects was low and, although there was a significant relationship with loprazolam treatment, this was not dose-related. Performance tests for residual effects (critical flicker fusion threshold and choice reaction time) showed no difference between the placebo and active treatment groups, suggesting that loprazolam, even in a dose of 1.0 mg, is unlikely to produce hangover effects. It is concluded that the findings confirm that loprazolam is an effective and well-tolerated hypnotic in the elderly and that a 0.5 mg dose would be an appropriate initial dose for older patients.