The determination of specific keto-acids in blood and urine was derived by modifying existing procedures. Drawn blood was mixed with 5 volumes of cold trichloroacetic-acid or metaphosphoric-acid and centrifuged. Urine samples were collected in bottles containing hydrogen-sulfoxide, treated with Lloyd's reagent, and filtered. Samples were treated with 2,4-dinitrophenylhydrazine extracted with ethyl-acetate to determine total keto-acids or xylene, toluene, or benzene to determine pyruvic-acid, 10 percent of the alpha-ketoglutaric-acid, or 5 percent of the oxalacetic-acid. Blood was reacted with hydrazine and sodium-hydroxide to determine total hydrazones. Color interference of less than 1 percent indicated hexoses, glucuronic-acid, ascorbic-acid, kojic-acid; 1 to 10 percent indicated acetylpyruvic-acid, acetoacetic-acid, levulinic-acid; greater than 10 percent indicated formaldehyde, acetaldehyde, glyoxal, methylglyoxal, glyceraldehyde, acetone, glyoxylic-acid, oxalacetic-acid, and alpha-ketoglutaric-acid. Of the phenylhydrazines tested, 2,4-dinitrophenylhydrazine gave the best combination of photometric density and reaction time. Reactions with pyruvic-acid, formaldehyde, acetaldehyde, glyoxal, glyoxlic-acid, glyceraldehyde, methylglyoxal, acetoacetic-acid, acetone, and levulinic-acid took place within 5 minutes. Reactions with oxalacetic-acid and alpha-ketoglutaric-acid took about 20 minutes. Different keto-acids showed varying solubilities in benzene, toluene, xylene, chloroform, ethyl-ether, caprylic-alcohol, and ethyl-acetate. Ethyl-acetate was the most consistent. The hydrazones exhibited differences in light absorption that could be used to determine the keto-acid and its relative quantity by a ratio of the wavelengths 420 and 520 millimicrons. The author concludes that the single extraction method described yields identical results to complete extraction methods and has greater specificity for alpha-keto-acids
