1. Riboflavin deficiency at two levels of severity was produced in weanling rats by feeding deficient diets for 6 weeks and using neck collars to prevent coprophagy. The severity of deficiency was monitored by growth, liver flavin levels and the activation coefficient of erythrocyte glutathione oxidoreductase (NAD(P)H) (EC 126.96.36.199). Control groups, receiving the same diet with ample added riboflavin, were fed either ad lib., or were pair-fed with the deficient animals. 2. The hepatic flavoenzyme, methylenetetrahydrofolate reductase (NADPH) (EC 188.8.131.52), was very markedly affected by severe riboflavin deficiency and was significantly, but less markedly, affected by the intermediate level of deficiency. This reduction in activity was due primarily to the direct effect of the diminished supply of riboflavin, and occurred to only a small extent as a result of inanition, demonstrated by a moderate reduction in activity in the more severely food-restricted of the two pair-fed groups. Since the enzyme is assayed in the presence of its flavin cofactor, FAD, it clearly cannot be reactivated in vitro, as some other depleted flavoenzymes can. The discriminatory ability in distinguishing between severe and moderate riboflavin deficiency in vivo confers some potential advantages on this oxidoreductase as a possible index of riboflavin status. 3. The hepatic activity of another key folate-metabolizing enzyme, dihydrofolate reductase (EC 184.108.40.206), was not diminished by riboflavin deficiency in the present study. 4. The ratio, labelled 5-methyltetrahydrofolic acid:other labelled compounds derived from intraperitoneally injected pteroylglutamic acid in extracts of hepatic tissue was significantly reduced in the riboflavin-deficient groups, indicating the possibility of an effect of riboflavin deficiency on folate metabolism in vivo.