The authors examined 1) effects of nortriptyline (NT) on electroencephalographic (EEG) sleep measures in elderly patients with bereavement-related depression in remission under randomized, double-blind, placebo-controlled conditions, and 2) the effects of clinical remission on sleep after discontinuation of medication. Subjects were classified as responders to placebo (n=9) or NT (n=18) and had EEG sleep studies at three time-points: before treatment (T1), remitted on medication or placebo (T2), and remitted off medication or placebo (T3). As compared with placebo, NT was differentially associated with decreases in REM sleep time and percent and increases in REM sleep density (T2). No changes in EEG sleep measures occurred in placebo responders. REM sleep measures in NT responders reverted to T1levels after T3, with persistence of robust clinical remission and normal subjective sleep quality. These data suggest that NT alters REM sleep, but that EEG sleep characteristics in bereavement-related depression persist into remission.