Eight patients affected by Alzheimer's Presenile Dementia (AD) received acute administration of physostigmine individual optimal dose per os (n = 4) or subcutaneously (n = 4). The individual physostigmine dose was assessed by means of serum cholinesterase activity monitoring. The possible beneficial effects after treatment were evaluated by using two memory tests: Reys' 15 words and Digit Span from Wechsler memory scale. Although a slight behavioral activation was noted in all patients after treatment, the comparison between mean scores obtained by AD patients in mnesic tests before and after the acute physostigmine administration, with either therapeutic modality, failed to reach the level of statistical significance. Some implications of these disappointing results are briefly discussed.