One approach to the differentiation of Alzheimer's disease (AD) and major depressive disorder (MDD) has been the application of the P300 event related brain potential (ERP) to assess the fundamental cognitive processes involving stimulus evaluation (Polich 1992). Twelve AD patients (60.5 ± 9 years), 12 MDD patients (61 ± 12 years), and 12 healthy controls (HC; 65.8 ± 6 years) participated in three consecutive tasks comprising 300 tones each, with a 10 minute pause in between. Two different tones were used as stimuli: 1400 Hz tones served as targets with 15% probability, and 800 Hz tones as nontargets with 85% probability. The first task was to count the number of infrequent tones silently. The second task involved responding to infrequent stimuli by pressing a button with the right index finger. In the third, subjects were required to monitor the stimulus series without any specific overt or covert responses. EEG activity was recorded at 21 positions on the scalp, eye movements being corrected by a modified subtraction procedure. EEG trials involving target tones were averaged separately for each task and subject. Group differences were proofed statistically at fixed latencies. Using overall MANOVA to assess the P300 maximum amplitudes across all groups and tasks, we found significant differences (p < 0.05). Subsequent t tests showed that the third task could differentiate between MDD and AD as well as between MDD and HC. These latter findings suggest that ERP can facilitate the early differential diagnosis of cognitive and affective disorders in aging.