To test the effects of a hypnotic drug administered on a regular basis, six adults (four women and two men; mean age 37 years) who complained of transient or short-term insomnia, took zolpidem 10 mg at bedtime for 4 consecutive weeks. The period of active treatment, preceded by a baseline placebo night, ended with a 4-night gradual tapering phase followed by 3 nights of placebo administration. After adaptation to the sleep laboratory, all subjects underwent five polysomnographic recordings; baseline placebo (night 1); 1st, 7th and 28th nights of hypnotic medication (nights 2, 3, and 4 respectively); 3rd placebo night after complete tapering (night 5) and completed a morning visual analogue scale (VAS) for evaluating sleep quality. The sleep recordings were scored according to the conventional procedures (macrostructure) and to the cyclic alternating pattern (CAP) rules (microstructure). Data analysis was based on a repeated-measures analysis of variance integrated by post-hoc comparisons. At the macrostructural level, significant overall modifications (p