Neuropsychometric tests (for instance the Alzheimer's Disease Assessment Scale, ADAS) and the EEG are often used in the diagnostic procedure of dementia. The validity of the instruments is only poorly investigated. The study aimed to investigate the accuracy of the discrimination between healthy controls and patients with dementia of the Alzheimer type (DAT) by ADAS and EEG. Thirty-six patients with DAT and 44 healthy controls were included. In a discriminant analysis of the 21 ADAS items and 18 EEG parameters (6 frequency bands, 12 topographic parameters), 6 ADAS items turned out to discriminate both groups with 100% sensitivity and specificity (remembering instructions, depression, following commands, pacing, restlessness and word finding difficulties). Regarding EEG parameters, 4 (topography of beta- and delta-activity and amplitude of delta-activity) led to a sensitivity and specificity of over 90%. Thus, both methods demonstrated an excellent discrimination between healthy controls and DAT. The slightly higher discrimination with the ADAS may depend on its closer relation to clinical symptoms. However, the EEG measuring functional activity reached nearly the same result. Both methods provide complementary information. A combination of both methods in the diagnostic procedure to detect dementia is recommended. Copyright (C) 2000 S. Karger AG, Basel.