The aim of this study was to elucidate physiological processes that are involved in the homeostatic regulation of REM sleep. Adult rats were chronically instrumented with sleep-wake recording electrodes. Following post-surgical recovery, rats were habituated extensively for freely moving polygraphic recording conditions. On the first experimental recording day (baseline day, BLD), polygraphic signs of undisturbed sleep-wake activities were recorded for 4 h (between 11:00 AM and 3:00 PM). During the second experimental recording day (REM sleep deprivation day, RDD), rats were selectively deprived of REM sleep for the first 2 h and then allowed to have normal sleepwake for the following 2 h. The results demonstrated that during the first 2 h, compared to BLD, RDD recordings exhibited 87.80% less time in REM sleep and 16% more time in non-REM (NREM) sleep. The total percentages of wakefulness remained comparable between the BLD and RDD. During the RDD, the mean number of REM sleep episodes was much higher than in the BLD, indicating increased REM sleep drive. Electroencephalographic (EEG) power spectral analysis revealed that selective REM sleep deprivation increased delta power but decreased theta power during the residual REM sleep. During the last 2 h, after REM sleep deprivation, rats spent 51% more time in REM sleep compared to the BLD. Also during this period, the number of REM sleep episodes with the shortest (5-30 s) and longest (>120 s) duration increased during the RDD. These findings suggest that the REM sleep homeostatic process involves increased delta-and decreased thetafrequency wave activities in the cortical EEG.