Sixteen patients diagnosed clinically and in the laboratory as having Alzheimer's disease (AD) of late onset (>70 years of age) type, were given drugs twice a day for sixteen weeks, in a double blind design. Seven patients were given lecithin (62% phosphatidyl choline) and choline chloride and 4 patients placebo. Five patients were first treated with lecithin and choline chloride for four weeks, then L‐dopa combined with benserazide hydrochloride was added for four weeks and finally L‐tryptophan was also given for eight weeks (per day mean ± SEM: 26 ± 0.53 g, 53 ± 1.7 mmol, 0.71 ± 0.050 mmol, 4.1 ± 0.30 mmol). In addition, nicotinamide, sodium salicylate and benserazide hydrochloride were given (per day: 8.2 mmol, 9.4 mmol and mean ± SEM: 40 ± 2.8 mg, respectively). Side reactions were studied twice a week, behavior and chemical effects from blodd and urine samples and after 8 and 16 weeks of treatment. The three precursors, which were intended to increase cholinergic, dopaminergic and serotonergic activities in brain. Two of these precursors, lecithin and phosphatidyl choline, are found in ordinary food and the third, L‐dopa, is normally produced by in vivo oxidation of L‐tyrosine in food. Lecithin and choline chloride did not decrease food intake, but these substances combined with L‐tryptophan did so, as did L‐dopa. The lecithin and choline chloride treatment increased the fasting plasma levels of choline and low density lipoprotein cholesterol but did not change the urinary excretions of cortisol, homovanillic acid (HVA), 5‐hydroxyindoleacetic acid (5‐HIAA), norepinephrine and epineph‐rine. The addition of L‐dopa and L‐tryptophan increased the urinary excretions of HVA, and 5‐HIAA and probably also the norepinephrine excretion. This complete multisubstitution schedule decreased the urinary excretion of epinephrine, did not change cortisol excretion and seemed to increase the urinary excretion ratios of HVA to 3‐methoxy‐4‐hydroxy‐mandelic acid (HMMA) and to HMMA + 3‐methoxy‐4‐hydroxyphenylglycol (MHPG). These increases were, however, not significant There were no consistent effects on the cognitive functions or performance with thi