OBJECTIVE: To evaluate the contribution of quantitative electroencephalographic (qEEG) analyses in the diagnosis of Alzheimer's disease (AD). METHOD: Thirty-five patients from the Neurology Outpatients Clinic of PUC-Campinas, diagnosed with AD according to the NINCDS/ADRDA were evaluated, and compared with a control group consisting of 30 individuals with no cognitive deficit. The procedures consisted of clinical-neurological, cognitive and behavioral analyses and the qEEG (absolute power and coherence). RESULTS: The AD group presented greater absolute power values in the delta and theta bands, greater theta/alpha indices and less frontal alpha and beta coherence. Logistic multiple regression models were constructed and those only showing variations in the qEEG (frontal alpha coherence and left frontal absolute theta power) showed an accuracy classification (72.3%) below that obtained in the mini-mental state examination (93%). CONCLUSION: The study of coherence and power in the qEEG showed a relatively limited accuracy with respect to its application in routine clinical practice.