OBJECTIVE: We tested the hypotheses that human brain glycogen is mobilized during hypoglycemia and its content increases above normal levels ("supercompensates") after hypoglycemia. RESEARCH DESIGN AND METHODS: We utilized in vivo (13)C nuclear magnetic resonance spectroscopy in conjunction with intravenous infusions of [(13)C]glucose in healthy volunteers to measure brain glycogen metabolism during and after euglycemic and hypoglycemic clamps. RESULTS: After an overnight intravenous infusion of 99% enriched [1-(13)C]glucose to prelabel glycogen, the rate of label wash-out from [1-(13)C]glycogen was higher (0.12 +/- 0.05 vs. 0.03 +/- 0.06 micromol x g(-1) x h(-1), means +/- SD, P < 0.02, n = 5) during a 2-h hyperinsulinemic-hypoglycemic clamp (glucose concentration 57.2 +/- 9.7 mg/dl) than during a hyperinsulinemic-euglycemic clamp (95.3 +/- 3.3 mg/dl), indicating mobilization of glucose units from glycogen during moderate hypoglycemia. Five additional healthy volunteers received intravenous 25-50% enriched [1-(13)C]glucose over 22-54 h after undergoing hyperinsulinemic-euglycemic (glucose concentration 92.4 +/- 2.3 mg/dl) and hyperinsulinemic-hypoglycemic (52.9 +/- 4.8 mg/dl) clamps separated by at least 1 month. Levels of newly synthesized glycogen measured from 4 to 80 h were higher after hypoglycemia than after euglycemia (P