The aim of this study was to evaluate the effects of a 200-mg administration of caffeine on polysomnographic sleep variables and quantitative sleep electroencephalography (EEG) in 12 young (20-30 years) and 12 middle-aged (40-60 years) moderate caffeine consumers (one to three cups of coffee per day). All subjects were submitted to both a caffeine (200 mg) and placebo (lactose) condition in a double-blind cross-over design. The conditions were separated by 1 week. Compared with the placebo condition, the evening ingestion of caffeine lengthened sleep latency, reduced sleep efficiency, and decreased sleep duration and amount of stage 2 sleep in both age groups. Caffeine also reduced spectral power in delta frequencies in frontal, central and parietal brain areas, but not in prefrontal (PF) and occipital regions. Moreover, caffeine increased spectral power in beta frequencies in frontal and central brain areas in both age groups. A suppression of spectral power in the PF area in low delta frequencies (0.5-1.00 Hz) and a rise in spectral power in the parietal region in high alpha (10.00-12.00 Hz) and beta frequencies (17.00-21.00, 23.00-25.00, 27.00-29.00 Hz) occurred solely in middle-aged subjects. No such changes were noticeable in young subjects. Generally, caffeine produced similar effects in young and middle-aged subjects. Only a few frequency bins showed more effects of caffeine in middle-aged subjects compared with young subjects. Furthermore, sleep EEG results do not entirely support the hypothesis that caffeine fully mimics the effects of a reduction of homeostatic sleep propensity when following a normal sleep-wake cycle.