Quantitative EEG (qEEG) findings in Parkinson disease (PD) have been reported in only five previous studies. In these studies, the sample size was small and the distribution of qEEG changes was not estimated. This is the first qEEG evaluation not only employing multiple logistic regression analysis but also estimating the distribution of qEEG changes. The subjects comprised 45 PD patients without remarkable dementia and 40 age-adjusted normal controls. The lack of ischemic lesions in all subjects was confirmed by MRI. Absolute power values were measured for four frequency bands from delta to beta. The electrodes were divided into six, viz. frontal pole, frontal, central, parietal, temporal, and occipital locations. We calculated the spectral ratio, i.e., the sum of the power values in the alpha and beta waves divided by the sum of the values in the slow waves. The dependent variable was either PD or normal control; the independent variables were the spectral ratios, age, sex, and Mini-Mental State Examination score. The significant predictive variables in PD were the spectral ratios at all electrode locations except for the frontal pole (frontal location: P = 0.025, other locations: P < 0.01). PD presented diffuse slowing in the qEEG when compared with age-adjusted normal controls.