Post-traumatic stress disorder (PTSD) overlaps major depression (MD) clinically, but differs with respect to treatment response and some biological markers. Sleep disturbances represent core features of PTSD and are also common in MD. Rapid eye movement sleep (REM) has been postulated to be involved in the pathophysiology of PTSD, and REM abnormalities occur in MD. Twenty-five patients with combat-related PTSD, 16 men with a principal diagnosis of MD, and 10 asymptomatic male controls were compared by polysomnography (PSG) under medication and substance-free conditions. Data were obtained from recordings made after an accommodation night. One subject from each group was excluded for significant apnea or limb movements. Sleep efficiency was decreased in the PTSD group compared to the MD and control groups. REM density was comparably increased in PTSD and MD groups, while the amount of REM sleep was reduced in PTSD compared to MD groups. These sleep measures were not significantly associated with co-morbid depression, substance-use disorder histories, or subclinical sleep apnea or limb movements within the PTSD group. These findings support sleep maintenance being impaired in chronic PTSD patients. Increased REM density in PTSD patients was replicated and was comparable to increases in the MD group. Divergence of REM time between these clinical groups suggests the possibility of different underlying mechanisms.