The inborn metabolic error in phenylpyruvic oligophrenia consists of the inability to hydroxylate phenylalanine to tryrosine in any significant amount. Recent work has focused mainly on the investigation of this hydroxylating system, its purification and mode of action (1-5), the inhibitory effect of phenylalanine and its derivatives on enzymes (6-9), and the prevention of mental retardation by a diet low in phenylalanine (10,11).With this diet the free phenylalanine of plasma and total body fluid can be adjusted to any value that might be required to study the effect of phenylalanine concentration on metabolic processes in vivo. Bickis, Kennedy and Quastel (12), for example, demonstrated that phenylalanine, at concentrations similar to those observed in the blood of phenylketonuric patients, inhibited the enzymatic degradation of tyrosine in vitro. We, on the other hand (13), were unable to find inhibition of the breakdown of orally administered p-hydroxyphenylpyruvic acid in patients with plasma concentrations of phenylalanine as high as 50 mg per 100 ml. This discrepancy could be explained, for instance, by a diminished rate of p-hydroxyphenylpyruvic acid formation from tyrosine in the in vitro experiments, or by a lower intracellular phenylalanine concentration in our in vivo study than one would have expected from the plasma phenylalanine levels. Since the distribution coefficient of phenylalanine between tissue and plasma has not yet been determined in vivo, the intracellular phenylalanine concentrations in our experiments were not known. In order to evaluate properly the significance of metabolic experiments with phenylalanine, some estimate must be made of the distribution of the phenylalanine between plasma and tissue.Since patients with phenylpyruvic oligophrenia * This work was supported by Research Grants 2729 and 3961 from the National Institute of Mental Health. demonstrate a defined metabolic variation from the norm, they also provide the opportunity to obtain further information that is not readily available in normal subjects. For example, pool size determinations of amino acids have rarely been performed and have mostly proved to be unsatisfactory, bceause they are based on experiments with N'5-labeled amino acids and many assumptions have to be made [see Wu, Sendroy and Bishop (14,15) and Tschudy and co-workers (16)]. Experiments similar to ours with S35_ labeled methionine in lnonphenylketonuric individuals have been published by Maurer (17). To our knowledge no studies on human beings have been made, so far, with C14-labeled amino acids, except by Gutman and co-workers (18), who estimated the pool size of glycine. The method used by them, however, was entirely different from ours.The use of C14-labeled amino acids eliminates the necessity of assuming, as did Sprinson, Rittenberg and San Pietro (19,20), that the rate of nitrogen transfer is faster than other metabolic processes involved here. It will be shown later that we have to make a similar assumption; namely, that equilibration of the ...