We have assessed the effect of midazolam on sleep in a model of transi
ent insomnia in healthy adults using polysomnographic recordings. The
subjects were randomly assigned to one of three treatment groups (plac
ebo or midazolam 7.5 or 15 mg) and spent a single night in the sleep l
aboratory. Midazolam or placebo were given double-blind immediately be
fore turning off the lights. Midazolam 15 mg was effective in inducing
sleep, while 7.5 mg and 15 mg produced improvement in the maintenance
of sleep. Subjectively, sleep latency and the number of awakenings we
re reduced dose-dependently. Midazolam did not impair psychomotor perf
ormance on the morning after administration.