The dose range of estazolam for hypnotic effects was studied in seven men and eight women (mean age 30.3 +/- 8.6 years) who complained of insomnia and had polysomnographic evidence of disturbed sleep. Patients slept in the laboratory and were monitored using standard polysomnographic techniques. Four consecutive nights in the laboratory with placebo, which served as baseline and screening nights, were followed by five 2-night drug administration periods separated by 5-day drug washout periods spent at home. Each of the patients received a sequence of four doses (0.25, 0.5, 1.0, and 2.0 mg) of estazolam and placebo administered according to a Latin square design and in a double-blind manner. Estazolam significantly increased total sleep time and reduced time awake during sleep in a dose-dependent manner. Sleep latency parameters were reduced systematically with increasing doses of estazolam, but these effects on sleep latency were not statistically significant. Increasing doses of estazolam had systematic and statistically significant effects on sleep stages. Subjective estimations of sleep were consistent with the polysomnographic findings bur were not statistically significant.