This trial was performed to investigate the efficacy of pyritinol in the treatment of senile dementia. Initially, a total of 183 inpatients were screened for eligibility. Of 164 patients who met the inclusion criteria, 156 completed the trial. Allocation of the patients to the Senile Dementia of the Alzheimer Type group or the Multi-Infarct Dementia group was based on the Hachinski Ischemic Score, computed tomography scans and electroencephalographic (EEG) findings. In a 12-week double-blind treatment phase either 200 mg pyritinol dihydrochloride-monohydrate or placebo was given 3 times daily. Confirmatory statistics included item 2 of the Clinical Global Impression, the total score of the Short Cognitive Performance Test (Syndrom Kurz Test) and the factor 'cognitive disturbances' of the Sandoz Clinical Assessment Geriatric scale. In addition, data on tolerance, of EEG brain mapping and of a responder analysis were evaluated based on descriptive statistics. The therapeutic efficacy of pyritinol was clearly demonstrated by confirmatory analysis as the drug was statistically significantly superior to placebo in all 3 target variables. The clinical relevance of the outcome was underlined by the analysis of the descriptive variables and by the convergence found at the different observation levels. The EEG mapping demonstrated significant differences between placebo and pyritinol, with the latter decreasing slow and increasing fast alpha and beta activity, which reflects improvement of vigilance. Based on the results of this trial, it can be accepted that the therapeutic effect of pyritinol is superior to placebo in patients with mild to moderate dementia of both degenerative and vascular etiology.