In a parallel-group, placebo-controlled, polysomnographic study with randomization, the possible occurrence of rebound insomnia was evaluated in 24 patients suffering from moderate to severe chronic insomnia and receiving either triazolam 0.5 mg, zolpidem 10 mg, or placebo. Treatment duration was 27 nights, followed by three placebo-controlled withdrawal nights. Both drugs showed significant efficacy compared to placebo during the active treatment period. A trend toward tolerance was noted in the triazolam group but not in the zolpidem one. The increase in total sleep time in the zolpidem group was accompanied by an increase in the number of sleep cycles. When active treatment was discontinued, clear rebound insomnia was present in the triazolam group while it was not possible to observe any rebound in the placebo and zolpidem groups. Subjective feelings of the patients, which were assessed by means of visual analog scales, correlated well with polysomnographic data. Our findings tend to indicate that, even after long-term treatment, zolpidem does not induce rebound insomnia or daytime anxiety.