Electroencephalographic (EEG) sleep studies may help to identify persistent versus episodic biological characteristics of major depressive disorder. This report examines longitudinal EEG sleep studies in depressed patients treated with psychotherapy alone. Nineteen patients were studied during a symptomatic baseline period and again during early remission after treatment with interpersonal psychotherapy (IPT). EEG sleep findings at baseline were not markedly abnormal, but they were similar to those in other published studies of young adult outpatients. No changes were found in visually scored EEG sleep measures between depression and early remission. Automated measures of delta sleep and rapid eye movement (REM) activity showed small state-related changes, with delta activity increasing from baseline to remission, and automated REM measures decreasing. Strong baseline-remission correlations were noted for most sleep measures, including slow wave sleep, phasic REM activity, and automated delta EEG counts; measures of sleep continuity and tonic REM sleep were not strongly correlated. Consistent adaptation effects across nights were observed for sleep continuity and REM measures during each clinical phase. These findings support the hypothesis that most visually scored EEG sleep measures, as well as the sleep adaptation process, are stable through the acute episode of depression, at least into early symptomatic remission. They also suggest that finer-grained automated analyses of delta and REM activity may provide more sensitive tools for examining state-related changes.