The sleep apnea syndrome is often associated with the syndromes of daytime exhaustion and involuntary daytime sleeping fits. The cause is assumed to be fragmentary sleep resulting from night-time arousal. The central nervous activation reactions caused by apnea or hypopnea, respectively, and not the movement arousal determine the sleep structure. We have examined 10 male patients in the age range 40-55 years (48 +/- 6 SD) before and during the first 3 nights of CPAP therapy. Cardiorespiratory polysomnography was performed in all four nights. Sleep way analyzed visually and differentiation was made between respiratory (RA) and movement arousal (MA). All 10 patients had a pronounced sleep apnea syndrome. Deep and dream sleep were reduced, significantly more respiratory arousals occurred than movement arousals. The SWS latency was shortened in the first therapy night, the deep and dream sleep proportions increased and the RA decreased significantly. No further significant changes in the sleep parameters occurred during the second and third nights. We found that the number of apnea/hypopnea was not equal to the number of RA. When less arousal was recognized it was suggestive of a deficit of the diversion function while more RA was indicative of additional respiratory events, e.g. pharyngeal obstructions and hyperventilations which were not recognized as apnea or hypopnea. In addition to its role in the differential diagnosis of sleeping disorders, in particular sleep apnea, arousal differentiation is also an important criterion for estimating the efficiency of CPAP therapy.