A threonine-deficient amino acid mixture was developed by decreasing the level of threonine in a well-balanced amino acid mixture while measuring the response in daily gain and protein retention in young growing rats. The diet that contained the highest level of threonine and yet permitted further responses in growth and protein retention was used to define a threonine-deficient amino acid mixture. This mixture was used to provide three dietary levels of threonine (0.38, 0.43 or 0.48%) in combination with three levels of relative excess of all other amino acids (0, 25 or 50% relative excess). Food intake, weight gain and changes in carcass composition were measured in the 21-day study. When voluntary food intake is used as a covariate in the analysis of these data, the level of the threonine-deficient amino acid mixture is positively correlated with body weight, dry matter, crude protein and ash gains and negatively associated with lipid deposition. However, there were no significant effects of amino acid excess on any of the responses when differences in voluntary food intake are accounted for in the statistical analysis. These data demonstrate that some aspects of threonine imbalance (food intake and lipid deposition) are dependent on the dietary level of threonine. The major effects of threonine imbalance are due to decreased voluntary food intake rather than changes in the efficiency of use of ingested threonine.