To test theories that response to sleep deprivation in depression is the result of either stress reactions or down-regulation of hyperarousal, the early morning cortisol and beta-endorphin levels of depressed sleep deprivation responders and nonresponders before and after sleep deprivation were compared (areas under the curve of 8 blood samples between 7.30 and 10 a.m.). The beta-endorphin response was significantly different in responders and nonresponders, whereas all other comparisons remained nonsignificant. The results do not support theories that sleep deprivation acts as a stressor, but are not contradictory to the hyperarousal hypothesis of sleep deprivation.