A 6-month double-blind, randomized, placebo-controlled clinical trial preceded by a 3-week single-blind, washout/run-in placebo phase was performed in male and female patients, 55-85 years of age with a clinical diagnosis of mild to moderate multi-infarct dementia according to DSM-III to evaluate the therapeutic efficacy and safety of nicergoline 30 mg b.i.d. Primary endpoints for efficacy were the changes in the Sandoz Clinical Assessment Geriatric Scale (SCAG) and Mini-Mental State Examination (MMSE) scores at the end of the treatment with respect to baseline. Secondary endpoints were Clinical Global Impression, 3 subtests of the Weschsler Adult Intelligence Scale and Blessed A scale for activities of daily living, and all endpoints in 2-month intervals. A total of 252 patients were screened, 136 patients entered the double-blind phase and were evaluated as intent-to-treat (ITT) patients. Fifteen patients were excluded from the efficacy analyses of valid cases (VC) due to protocol violations or because they dropped out of the study prematurely. Confirmatory efficacy analysis after 6 months of treatment revealed superiority of nicergoline treatment with p < 0.01 for both SCAG and MMSE scores (ITT and VC). Subsequent descriptive efficacy analysis resulted in significant differences in favor of nicergoline, in the majority of cases as early as 2 months after start of treatment. Nicergoline was well tolerated and a similar number of adverse events were observed in both the placebo and the nicergoline group.