An amino acid mixture devoid of tryptophan, given orally, was previously shown to reduce cerebrospinal fluid levels of tryptophan and 5-hydroxyindoleacetic acid in vervet monkeys, as compared to a control mixture containing all essential amino acids. In the present study, we tested the possibility that a similar amino acid mixture containing tryptophan, but devoid of phenylalanine and tyrosine (the amino acid precursors of catecholamine neurotransmitters), would influence dopamine and noradrenaline metabolism. Five hours after the administration of this mixture to vervet monkeys, cerebrospinal fluid levels of homovanillic acid and 3-methoxy-4-hydroxyphenylethylene glycol were reduced by 27.4% and 26.9%, respectively. Both effects were statistically significant. Plasma tyrosine (-30%) and the ratio of tyrosine to the sum of other large neutral amino acids (sigmaLNAA) were also significantly reduced. The behavioral efficacy of phenylalanine/tyrosine depletion was compared with that of tryptophan depletion in a primate model of voluntary alcohol consumption. All three drinks lowered alcohol consumption, but the effects of the tryptophan-deficient amino acid mixture were not different from those of the balanced amino acid control. The phenylalanine/tyrosine-deficient drink differentially lowered alcohol consumption, consistent with other data in this species and elsewhere implicating dopamine in the rewarding effects of alcohol.