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 < 0.035) emerged only from slow-wave sleep (SWS). The amounts of SWS were enhanced along the entire drug period, but a significant difference was found only between the baseline night (10%) and the first night of drug administration (20%). At the microstructural level, CAP rate (the ratio of CAP time to non-rapid eye movement sleep time) showed overall significant variations throughout the entire protocol period (p < 0.0001). Compared with baseline night 1 (59%), the CAP rate was significantly lower on drug nights 2 (32%), 3 (37%), and 4 (38%). The increase in the CAP rate found on night 5 (43%) was still significantly below the baseline value. The VAS scores showed significant overall changes (p < 0.0001), with improved values during the active treatment period.