O ther investigators have shown that fructose infusion in normal man and rats acutely depletes hepatic ATP and P_i and increases the rate of uric acid formation by the degradation of preformed nucleotides. We postulated that a similar mechanism of ATP depletion might be present in patients with glucose-6-phosphatase deficiency (GSD-I) as a result of ATP consumption during glycogenolysis and resulting excess glycolysis. The postulate was tested by measurement of: (a) hepatic content of ATP, glycogen, phosphorylated sugars, and phosphorylase activities before and after increasing glycolysis by glucagon infusion and (b) plasma urate levels and urate excretion before and after therapy designed to maintain blood glucose levels above 70 mg/dl and thus prevent excess glycogenolysis and glycolysis.Glucagon infusion in seven patients with GSD-I caused a decrease in hepatic ATP from 2.25 ± 0.09 to 0.73 ± 0.06 μmol/g liver (P <0.01), within 5 min, persisting in one patient to 20 min (1.3 μmol/g). Three patients with GSD other than GSD-I (controls), and 10 normal rats, showed no change in ATP levels after glucagon infusion. Glucagon caused an increase in hepatic phosphorylase activity from 163 ± 21 to 311 ± 17 μmol/min per g protein (P <0.01), and a decrease in glycogen content from 8.96 ± 0.51 to 6.68 ± 0.38% weight (P <0.01). Hepatic content of phosphorylated hexoses measured in two patients, showed the following mean increases in response to glucagon; glucose-6-phosphate (from 0.25 to 0.98 μmol/g liver), fructose-6-phosphate (from 0.17 to 0.45 μmol/g liver), and fructose-1,6-diphosphate (from 0.09 to 1.28 μmol/g) within 5 min. These changes, except for glucose-6-phosphate, returned toward preinfusion levels within 20 min.Treatment consisted of continuous intragastric feedings of a high glucose dietary mixture. Such treatment increased blood glucose from a mean level of 62 (range 28-96) to 86 (range 71-143) mg/dl (P <0.02), decreased plasma glucagon from a mean of 190 (range 171-208) to 56 (range 30-70) pg/ml (P <0.01), but caused no significant change in insulin levels. Urate output measured in three patients showed an initial increase, coinciding with a decrease in plasma lactate and triglyceride levels, then decreased to normal within 3 days after treatment. Normalization of urate excretion was associated with normalization of serum uric acid.We suggest that the maintenance of blood glucose levels above 70 mg/dl is effective in reducing serum urate levels and that transient and recurrent depletion of hepatic ATP due to glycogenolysis is contributory in the genesis of hyperuricemia in untreated patients with GSD-I.