In a controlled double-blind study, 78 patients with moderate senile dementia were randomly assigned to 3-months' treatment with 200 mg naftidrofuryl twice daily in slow-release form or with placebo. The patients were selected on the basis of the EACG (Echelle d'Appréciation Clinique en Gériatrie), which was also used as main criterion for confirmatory analysis of the efficacy of naftidrofuryl. A further main criterion was the SGRS (Stockton Geriatric Rating Scale). Secondary criteria examined were visual memory as tested in the Benton test, verbal memory as tested in the Rey test and numerical memory as tested in the digit-span test. The patients' concentration was tested in the Zazzo test. The tests were performed before the patients were admitted to the study and after 1 and 3-months' treatment. Confirmatory analysis of the primary criteria showed a significant treatment difference in favour of naftidrofuryl in the form of improvement in the global symptoms of senile dementia both in the EACG and in the SGRS, the scores in the naftidrofuryl group improving by 15% compared with only 5% in the placebo group. This was paralleled by a significant improvement in visual and verbal memory in the naftidrofuryl group in comparison with placebo. The naftidrofuryl patients also showed a greater improvement in concentration than the placebo group. One patient in the naftidrofuryl group suffered briefly from gastro-intestinal symptoms. There were no changes in routine laboratory parameters studied.