It was hypothesized that the metabolic effects of caffeine, which can be objectively measured (i.e. physiological, "arousal"), could be used to develop a physiological arousal model of chronic insomnia in a group of normal young adults. Twelve normal young adult males participated for 11 nights after laboratory adaptation. Subjects received 400 mg of caffeine three times a day for 7 nights and days. As predicted, the use of caffeine resulted in increased metabolic rate. Sleep efficiency was significantly reduced by caffeine and multiple sleep latency tests (MSLTs) were significantly increased. Some adaptation to the metabolic, sleep efficiency, and MSLT effects of caffeine was seen over the week of administration. Withdrawal effects (i.e. rebound sleep or sleepiness) were not seen for metabolic, MSLT or sleep variables. The data indicated that caffeine was effective in producing significant metabolic and sleep effects and that those effects were related. The results were consistent with the interpretation that a chronic decrease in sleep efficiency associated with increased physiological arousal, although producing subjective dysphoria, does not produce a physiological sleep debt.