Chanda, McNaught & Owen(i) found that injection of thyroxine increased the ratio of bound to free thiamine (vitamin B x ) in the milk of the cow, and Thompson (2) found that ingestion of iodinated casein had the same effect. Injection of thiouracil(i) decreased this ratio, but neither thyroxine nor thiouracil affected the total output of thiamine in the milk. The increase of the bound form engendered by thyroxine was due partly to an increase in thiamine pyrophosphate (cocarboxylase) and partly to an increase in proteinbound thiamine. The riboflavin of cow's milk is partly free and partly combined in a non-ultrafiltrable form (3). The bound riboflavin is in the form of flavinadenine dinucleotide (FAD) for when set free, presumably from protein, by trichloroacetic acid, it behaves like FAD on chromotography or electrophoresis. The following experiments were done to see whether injection of thyroxine affects the partition of riboflavin between its free and its bound forms in cow's milk. EXPERIMENTAL Six Ayrshire cows, 1 month calved, were used. The experiment was divided into Period 1 of 20 days, Period 2 of 14 days and Period 3 of 20 days. During Period 2 three of the cows each received subcutaneously 6-2 mg. L-thyroxine (sodium salt) daily. Total and bound riboflavin were determined in the milk. Milk yields and alkaline phosphatase in the milk and the heart rate of each cow were recorded to show whether the thyroxine was producing its expected physiological effects.Heart rate was recorded at the time of afternoon milking each day. Alkaline phosphatase was estimated by the method of Kay, Aschaffenburg & Neave (4), and total riboflavin in milk by the method of Bessey, Lowry & Love (5) in which the fluorescence of the riboflavin was measured directly in the supernatant liquid from a precipitation of the milk with trichloroacetic acid. After neutralization with K 2 HP0 4 , the fluorescence of the solution was measured in a laboratory-built fluorimeter of the Lowry (6) type as described by Crossland, Owen & Proudfoot(7). To estimate the partition of riboflavin between the free and bound forms a chromatographic separation was made. The milk was treated in the cold with trichloroacetic acid to liberate FAD from its combination with protein.The precipitate was rejected, and the remaining liquid was neutralized with K 2 HPO 4 and extracted with phenol. Addition of ether to the phenol threw out an aqueous phase containing all the riboflavin and the FAD. Residual ether in the separated aqueous phase was evaporated by warming in hot water. Portions of the aqueous extract were chromatographed on Whatman no. 31 paper as described by Crossland et al(l). The riboflavin spot was then eluted and measured fluorimetrically. The difference between the total and that eluted from the chromatogram gave a measure of riboflavin bound to protein as FAD.