Diabetes and fasting provoke an increase in heart glycogen content, despite a decline in the amount of active glycogen synthase present. To determine if the activity of glycogen synthase i is still rate limiting for glycogen synthesis, we used 13C-nuclear magnetic resonance to measure the in vivo rate of glycogen synthesis and compared this with the activity of glycogen synthase and phosphorylase measured in tissue extracts using physiological concentrations of substrates and activators. In the basal state the activity of glycogen synthase i was depressed in the diabetic and fasted hearts (P less than 0.01). The rate of heart glycogen synthesis was measured during a 50-min infusion of D-[1-13C]-glucose (10 mg/min) and insulin (1 U/min) and averaged 0.32 +/- 0.04 mumol.min-1.g wet wt-1 in controls and was diminished in both the diabetic (0.18 +/- 0.04 mumol.min-1.g wet wt-1) and fasted (0.16 +/- 0.03 mumol.min-1.g wet wt-1) and fasted (0.16 +/- 0.03 mumol.min-1.g wet wt-1) rats (P less than 0.05 for each). During the glucose and insulin infusion the average activity of glycogen synthase i was greater in control than diabetic or fasted hearts (P less than 0.01 for each) and approximated the rates of net glycogen synthesis in each group. In contrast, there were no significant differences in phosphorylase alpha activity, measured in tissue extracts, among the three groups. Furthermore, although this phosphorylase alpha activity greatly exceeded synthase activity, it did not appear to be expressed in vivo. We conclude that in normal, diabetic, and fasted rats, glycogen synthase is rate limiting for glycogen synthesis.