There is a great deal of knowledge about the value of biochemical indices in the severe states of PEM (kwashiorkor or marasImus). Not very much information in this regard is available in marginal or moderate malnutrition, but from the point of view of public health, moderate malnutrition presents a far more wide-spread problem than kwashiorkor or marasmus. In experimental models of protein-energy malnutrition, animals have been subjected to rather severe deficiencies of protein or energy or both and very little is known about marginal protein malnutrition produced experimentally. The work reported here was an attempt to assess the value of biochemical methods in marginal protein deficiency by using a more appropriate model. One other important consideration was that the measurements to be used in the field in developing countries must be simple, involving the minimum of equipment and of samples to be analysed. The measurements should also be sensitive to small changes in dietary- protein intake. The work was undertaken between October, 1971 and October, 1974, Groups of rats weighing on average 71 g were fed for five weeks diets containing per kg 80, 120, 150, 180 and 240 g of casein, all of which allowed for substantial growth. Measurements were made at weekly intervals of body weight, serum total protein, albumin, transferrin, cholinesterase, ribonuclease and creatine kinase. The sum of the three branched chain amino acids was determined by thin-layer chromatography in 5 ul of serum. The recovery was checked with radioactive valine and averaged 85%. Not all measurements were made on all groups of rats. In some experiments muscle mass was measured at the end of the five weeks' feeding period. The biochemical teats described could be made on as little as 50 ul of serum. The most sensitive indicators apart from the body weight were serum albumin and the branched chain amino acids. With the amino acid test, two groups of rats on 12% and 18% casein diets could be distinguished. With the albumin test the groups showed differences which were significant at the level of 5%. It was clear from the results that the greater the number of samples analysed the more sensitive the two methods become in differentiating rats on the various regimes. The ease and simplicity with which the two methods could be used and their importance in the investigation of marginal protein deficiency is discussed in relation to field studies in developing countries where such disease abounds.