Diphenylamine (DPA), mixed with an adequate diet in concentrations of 0.01, 0.10, and 1.0%, was fed daily to male and female beagles for two years. Growth of both sexes was arrested by the 0.1 and 1.0% diets, the total two-year DPA intakes being, respectively, 0.14 and approximately 1.4 kg per animal. A pronounced anemia developed in the dogs on the 1.0% diet, one of mild degree appearing at the 0.1% level. Leukocyte numbers and differential percentages remained normal. After two years the erythrocytes of the dogs on the 1.0% diet showed a moderately decreased resistance to hypotonicity. Sulfobromophthalein tests of liver function from day 618 to day 627 indicated a moderate degree of liver damage in the dogs fed the 1.0% diet. Blood glucose concentrations were all within normal limits. Phenolsulfonephthalein tests of kidney function gave normal values, and the urines of the animals were negative for albumin and glucose. Twenty-four-hour urine collections between days 692 and 732 from individual dogs on controlled water intakes showed no significant difference between the excretion rates of the controls and those on the 1.0% DPA diet. All the urine specific gravities were within normal limits. All lesions discovered were in the dogs fed the 1.0% diet, these being a peripherolobular fatty change in the liver with a marked increase in liver weight and ether-extractable lipids; a mild hemosiderosis of the spleen, kidneys, and bone marrow; and a slight increase in kidney weight. If it be assumed that high-purity DPA, like that used in the foregoing experiments, has a toxicity threshold for man similar to that for dogs, any health hazard to the consumer of apples protected from scald by the proper use of such DPA should be extremely remote.