Ten patients diagnosed clinically as having Alzheimer's disease (age range 54 to 73 years) were given 35 g/day of a 53% lecithin mixture for two weeks, in a double-blind crossover design. Patients received memory training during the lecithin condition and "placebo training" during the placebo drug condition. Repeated assessment with Buschke's Selective Reminding Procedure provided no evidence of a therapeutic lecithin effect either during the two week clinical trial or during longer follow-up trials. Noninvasive measurement of regional cerebral blood flow (nr CBF) by the 133Xe-inhalation method, and repeated EEG's also failed to demonstrate a therapeutic effect. These latter indices reflected a decline in cerebral function over the course of the study, a finding which paralleled clinical impressions in four of the patients. Follow-up trials of memory training under placebo and lecithin conditions provided some suggestion that memory training may lead to some immediate improvement in list-learning ability, but the improvement was not well maintained overtime.