Summary: The application of automated analysis in the measurement of sleep electroencephalogram delta activity allows a more precise temporal description of slow wave sleep changes in normal and pathological aging than do standard, all-night, slow wave sleep measures. Thus, with a baseline crossing technique, elderly depressives were shown to have a higher delta count per minute during the second NREM sleep period than did controls or demented subjects. This difference reflected greater activity per minute in the 2-3 Hz frequency band (75-200 fLY). Key Words: Period and amplitude analysis-Slow wave sleep-Elderly depression-Alzheimer dementia.The nocturnal sleep of elderly depressed and Alzheimer dementia patients can be distinguished by differences in REM sleep and sleep maintenance indices (1,2). Other studies have also reported decreased slow wave sleep time and percentage of time spent asleep in Alzheimer patients (particularly the more severely demented) as compared with controls (3,4), as well as decreased numbers of sleep spindles and K-complexes correlating with severity of dementia (2-5). In general, however, the standard all-night summary measures of slow wave sleep (minutes, percent of time asleep) are limited in their ability to demonstrate group differences among healthy elderly controls, patients with probable Alzheimer dementia, and elderly depressive patients. This may reflect the age-dependent decrease in slow wave sleep (6-8). Moreover, all-night summary indices convey no information about differences in the intranight temporal distribution of slow wave activity in the sleep electroencephalogram (EEG) or differences in the frequency composition of slow wave sleep (9). We report here the application of a baseline crossing technique with an amplitude criterion in a comparative EEG sleep study of elderly depressed patients, probable Alzheimer patients, and healthy elderly controls.