L-[U-14C]Threonine was infused at a steady rate to non-anaesthetized rats starved for 1 or 3 days and to diabetic rats starved for 1 day. The rates of turnover of threonine, calculated from the equilibrium specific radioactivity (SA) of plasma threonine, were 5.79 +/- 1.00, 11.67 +/- 1.43 and 13.35 +/- 1.85 mumol/min per kg body wt. in 1-day-starved, 3-day-starved and diabetic rats respectively. The calculated turnover rate of threonine agreed well with the rate expected from the rate of protein turnover reported in the literature. The equilibrium SA of plasma alanine was 5.1-9.8% of that of threonine in the three groups of rats. The equilibrium SA of glucose was 1.42 and 2.90% of that of threonine in 1-day- and 3-day-starved rats respectively. From the non-equilibrium SA of glucose, it is estimated that a higher percentage of 14C atoms is transferred from threonine to glucose in diabetic than in non-diabetic rats. In spite of increases in gluconeogenesis from threonine in long-starved or diabetic rats, we conclude that threonine remains a minor contributor to plasma glucose. Since it is an essential amino acid, its turnover and contribution to the formation of plasma glucose is an index of catabolism and gluconeogenesis from tissue protein.