The comparison of the sleep pattern of aged, long-term schizophrenic patients never biologically treated with an age-matched group of normal controls has shown that significant differences were present only in the mean number of REM periods, mean duration of each REM period, mean REM latency, the number of awakenings during the REM period or immediately after it, and mean sleep latency between both groups. The role of several factors that could account for these significant differences is discussed and an increased REM-period vulnerability in long-term schizophrenic patients is hypothesized.