By considering the prevalence of depression in subjects with sleep disorders, our purpose is diagnosis of depression in these subjects. Shortened REM latency (i.e, the interval between sleep onset and the occurrence of the first REM period), increased REM sleep duration and increased REM density (i.e., the frequency of rapid eye movements per REM period) have been considered as biological markers of depression, so to diagnose of depression in subjects with sleep disorders we used analysis of EEG signal in REM sleep. The features of frequency and time domains in four channels (C3-C4-O1-O2) and four bands (delta, alpha, beta ,gamma) of REM sleep EEG were compared in 10 depressed (37.5 ± 8.34 years; mean ± standard deviation) and 14 non-depressed (36.6 ± 11.1 years; mean ± standard deviation) subjects with sleep disorders. According to results, most of the significant differences (P<0.05) have related to C3 and O1, in first stage and first period of REM sleep. Our results also show significant increase of beta activity in depressed subjects. Mahalanobis distance has classified 91.7% of the depressed and non-depressed subjects correctly.