This article explored the relationship between daytime affect and REM sleep in 45 depressed men before and after treatment with cognitive-behavioral therapy and in a control group of 43 healthy subjects. The intensity of daytime affect (as measured by the sum of positive and negative affects) in depressed men correlated significantly and positively with phasic REM sleep measures at both pre- and posttreatment. This relationship was not found in healthy control subjects. In depressed men, both affect intensity and phasic REM sleep measures decreased over the course of treatment. The results suggest a relationship between phasic REM sleep and intensity of affect reported by depressed men. On the basis of this preliminary observation, it was hypothesized that abnormalities in phasic REM sleep in depressed patients are related, in part, to fundamental alterations in the intensity of their affective experience.