Previous studies have shown that the activity of serotonin-containing raphe neurons in cats is almost completely suppressed during rapid eye movement (REM) sleep. However, since raphe unit activity is known to be grossly correlated with the level of behavioral arousal or tonic motor activity, this decrease in activity during REM sleep may be simply due to the fact that tonic EMG activity or motoric output is at a minimum. On the other hand, raphe unit activity may be related to the state (i.e. REM sleep) of the organism. To test these competing hypotheses, in the present study we compared raphe unit activity in normal cats with that in cats that display REM sleep without atonia (produced by bilateral lesions of the pontine tegmentum). These lesioned cats manifest episodes which, by all criteria, appear to be REM sleep except that they display overt behavior, presumably because the mechanism normally responsible for producing atonia has been disrupted. Although the activity of raphe neurons in lesioned cats during REM sleep without atonia was significantly below that seen in these cats during waking, the level of activity was often impressive. This is especially true when those animals that displayed the greatest degree of tonic motor activity during REM sleep (group IV animals) are considered separately. In these cats, the depression was only 40.5% below their quiet waking level, whereas in lesioned cats displaying less tonic motor activity (Group II animals), raphe discharge rate was 65.6% below their quiet waking level. The discharge rate of raphe neurons during REM sleep in lesioned cats was more than 6-fold greater than that seen in normal animals. These data, in conjunction with other recent results from our laboratory, suggest that the decrease in raphe unit activity during REM sleep is largely a concomitant of the atonia which characterizes that state. These data are discussed within the general context of the relationship between raphe unit discharge and the activity of central motor systems.