This multicentre, double-blind, randomised, placebo-controlled, parallel-group study compared the subjective hypnotic efficacy of nightly doses of zolpidem 10 mg or zolpidem 15 mg with that of placebo in patients with chronic insomnia. Patients received placebo or one of the zolpidem doses for 31 consecutive nights, followed by 4 nights of single-blind placebo administration. Self-reported subjective sleep latency (SSL), subjective total sleep time (SST) and other sleep-related parameters were collected via sleep questionnaires. SSL was significantly shorter in the zolpidem 10 mg and 15 mg groups than in the placebo group throughout active treatment. There was no significant difference between the two zolpidem groups. During the first week of double-blind treatment, SST was significantly longer in both zolpidem groups than in the placebo group, and remained numerically greater throughout active treatment. After discontinuation of zolpidem 10 mg, SSL and SST were not significantly different from placebo at any time. Furthermore, SSL was numerically shorter and SST was numerically longer than at baseline. Adverse-event incidence rates were 43, 57 and 70% for the placebo, zolpidem 10 mg and zolpidem 15 mg groups, respectively. The most common events that occurred with greater frequency in the zolpidem groups than in the placebo group were drowsiness, lethargy and dizziness. It was concluded that administration of nightly doses of zolpidem 10 mg to patients with chronic insomnia for 31 nights resulted in a continuous and significant reduction in SSL and in a persistent increase in SST, without the induction of tolerance during, or rebound insomnia after, this period.