Ever since electroencephalographic research began, interest has been focused on problems of the physiology, pathology, and symptomatology of the relationship between sleep and wakefulness. Following preliminaries, the present paper describes a combined procedure for the determination of the EEG preferential frequency, and by random sampling statistically examines its suitability for indicating the fluctuations of activation level as a control measure. A trained volunteer gradually increased his degree of wakefulness, reliably producing parallel changes in the preferential frequency in three frequency bands. A standardised manual period analysis as control produced largely the same results. The preferential frequency has considerable value for indicating the degree of wakefulness.