I. In three separate experiments, four groups of five to eight young male rats were fed either (i) a high-protein diet, for which the net dietary protein:total metabolizable energy ratio (NDp:E) was 0-1 (HP diet); or (ii) a low-protein diet, for which NDp:E was 0-04 (LP diet). In both these groups, food intake was ad lib. In group (iii) the HP diet was given in an amount approximately equal to that taken by the LP group fed ad lib. (HP-restricted). In group (iv) rats were fasted for 48 h after receiving the HP diet (HP-fasted). Each experiment lasted 4 weeks. 2. In the LP and HP-restricted groups, food intake was about 50% of that of the HP rats, while body-weight, after 4 weeks on diet was about 35% and 55% of that of HP rats, for LP and HP-restricted respectively. Both groups of malnourished rats gained some weight during the experiment. 3. Measurements of oral glucose tolerance and plasma insulin levels were made in the fourth week. LP and HP-restricted rats both showed low fasting insulin levels and low insulin to glucose ratios during the glucose tolerance tests; the LP rats were more seriously affected. 4. At the end of the fourth week the rats were killed and blood, liver and gastrocnemius muscle were analysed. LP rats showed specifically and consistently low values for haemoglobin and plasma protein concentration, and low activities of hepatic glucose-6-phosphatase (EC 3-1-3-9) and of alanine aminotransferase (EC 2.6.1.2) in liver and muscle. The activity of hepatic aspartate aminotransferase (EC 2.6.1.1) was, if anything, increased. The plasma amino acid concentrations and ratios showed a specific fall in branched-chain amino acids. Liver fat concentration was consistently elevated. The HP-restricted rats had normal values for haemoglobin, plasma protein andliver fat, and near-normal values for plasma amino acids. Hepatic alanine aminotransferase showed increased activity compared with HP rats, but muscle alanine aminotransferase showed reduced activity. The HP-fasted rats had increased haemoglobin, plasma protein and liver fat concentration, and very low liver glycogen concentrations. Hepatic alanine aminotransferase activity was elevated. Plasma alanine concentration was specifically reduced. 5. The results are consistent with suppression of gluconeogenesis, liver dysfunction and essential amino acid deprivation in LP rats. These biochemical changes found in rats on a low intake of a diet of low protein and high carbohydrate value are similar to those found in kwashiorkor. An equally low intake of a diet of good protein value (HP-restricted) led to marginally better growth, accompanied by biochemical signs of increased gluconeogenesis, analogous to those reported for nutritional marasmus. This nutritional state was not biochemically identical with that of acute fasting. 6. The results are discussed in terms of the consistency of the rat model, and its contribution to understanding biochemical changes found in infant malnutrition.