Myoclonic twitching is a ubiquitous feature of infant behavior that has been used as an index of active sleep. Although the active sleep of infants differs in some ways from the REM sleep of adults, their marked similarities have led many to view them them as homologous behavioral states. Recently, however, this view has been challenged. One avenue for resolving this issue entails examination of the neural substrates of active sleep. If the neural substrates of active sleep were found to be similar to those of REM sleep, then this would support the view that the two states are homologous. Therefore, in the present study, decerebrations were performed in the pons and midbrain to determine whether the mesopontine region is important for the expression of active sleep in infants, just as it is for the expression of REM sleep in adults. It was found that, in comparison to controls, caudal pontine decerebrations reduced myoclonic twitching by 76%, rostral pontine decerebrations reduced twitching by 40%, and midbrain transections had no significant effect on twitching. Moreover, analysis of the temporal organization of twitching indicated that pontine decerebrations predominantly affected high-frequency twitching while leaving unaffected the low-frequency twitching that is thought to be contributed by local spinal circuits at this age. These results indicate that the mesopontine region plays a central role in the expression of active sleep in infant rats.