The method of preparation of a Milk Substitute, formula 5410, and its nutritive value on rats have been reported in previous communications. Further tests of its nutritive effect were made by feeding a group of 100 infants including 4 premature and 10 undernourished. Another group of 43 infants fed with human breast milk or cow's milk were observed during the same period as controls. Among the infants of the two groups there were 6 pairs of twins; one of each pair was fed with the Milk Substitute and the other with human or cow's milk. Great majority of the infants were cared by their own mothers whose cooperation had been secured, while the rest of the infants (10%) were kept in our nursery.The Milk Substitute feeding was started with infants between half and six months of age, with an average age of three months. Observations were continued until they all reached one year old. Fish liver oil and ascorbic acid were supplied to infants of both the experimental and the control groups. Vegetable soup and fruit juice were given to the infants before five months of age. Then various kinds of other foods, including rice porridge, wheat flour nudols, bread, crackers, mashed meat and vegetable and egg yolk were individually supplemented furnishing not more than 1/3 of the total caloric intake, (at the age of one year), as the family income permitted.The appetite of the experimental infants was normal. There was no digestive disturbances and abnormal frequency and appearance of the stool. The average daily caloric intake was similar to those recorded in the literature.The body weight and height of the babies in the experimental group fell well within the normal range. The average initial body weight and height of the infants in the experimental group were somewhat lower than those of the control; however, the height reached .that of the control at the age of 7 months, and the weight at about 11 months. The red blood cell counts and the hemoglobin value of the blood were normal.Values of blood alkaline phosphatase activity higher than 15 Bodansky units were shown in 20 babies of the experimental group (27%) and 8 babies of the control group (23%) at the beginning of the observation; only 4 babies in the experimental group were found abnormal high blood alkaline phosphatase value at the end. At the beginning, 3 cases of rickets and 3 cases of suspected rickets were indicated by the radiograph of the wrist bones of the babies in the experimental group, and only one suspected ricket in the control group were observed; and they all became normal at the end of the experiment. Besides, there were no other signs of any nutritional deficiency among all the infants.