We used a clinical pharmacological model to test pyritinol versus placebo in patients with mental deficiency and a clinical diagnosis of beginning chronic brain syndrome. Following a two weeks' washout phase, 50 patients were randomly allocated to two treatment groups of 25 patients each, receiving either 200 mg pyritinol three times daily, or placebo under double-blind conditions. The treatment period lasted 8 weeks. To be included in the study, patients had to have at least 50% subvigil phases in the 15-min EEG resting recording. We define such behaviour as a neurophysiological disturbance of vigilance. Scores in the Benton test were to be 2 points below the expected value, and/or the NAF score was to be above a standard value attained in an old peoples' home (greater than or equal to 14). We used this clinical pharmacological model for an internal validation of our Vigilance Index (VI). According to our definition, the Vigilance Index should express vigilance in the sense of an optimalization of the neuronal system to enable this system to perform better. The delta F power and the alpha slow-wave index have been considered as vigilance-indicative variables in the EEG. We believe that vigilance can be better expressed by a multidimensional approach, which takes into account all EEG elements that express vigilance, such as the replacement of the occipital basic rhythm (e.g. alpha or beta rhythm) into slow waves, the lowering of the dominant occipital frequency (be it an alpha or beta frequency), the anteriorization of the basic rhythm in the occipital field to the frontal region.(ABSTRACT TRUNCATED AT 250 WORDS)