In the course of studies of the urinary excretion of creatinine by premature infants, a number of the infants fed cow's milk were found to excrete a substance which gave the Waffe reaction but which was not creatinine. The interfering chromogen was differentiated from creatinine by its solubility in ether, its lack of adsorption by Lloyd's reagent and its partial destruction by heat (1000 C.) in acid solution. It also gave a yellow color with dinitrobenzoic acid in alkaline solution instead of the purple-rose color given by creatinine (1). Acetone and acetoacetic acid were not present in the urines. This chance discovery initiated the present studies.The fact that the chromogen appeared only in the urine of infants receiving cow's milk diets relatively high in protein (5 grams or more per kgm.) suggested that it might be an intermediary product of protein metabolism. Qualitative and quantitative (2) tests of chromogenic urines revealed usual amounts of amino acids. Qualitative tests for tryptophane and the indole ring were negative. The Millon test for the hydroxyphenyl group was strongly positive, indicating that the intermediary products in the urine were probably derived from the metabolism of tyrosine and possibly of phenylalanine, an hypothesis which was substantiated by feeding observations with pure amino acids. The absence of darkening of the urine during aeration excluded the presence of significant amounts of homogentisic acid (alkaptonuria).Isolation and identification of intermediary metabolites