Experiments in adult animals have indicated that hyperglycemia accentuates whereas hypoglycemia ameliorates hypoxic-ischemic brain damage. To determine whether hypoglycemia is protective or deleterious to the perinatal brain subjected to hypoxia-ischemia, 7-d postnatal rats were rendered hypoglycemic either by receiving an s.c. injection of insulin or fasting for 12 h. All rat pups underwent unilateral common carotid artery ligation followed by exposure to 8% oxygen-balance nitrogen at 37 degrees C for 2 h. Control animals (no insulin or fasting) received s.c. injections of normal saline. Mean blood glucose concentrations were 5.4 +/- 0.1, 4.3 +/- 0.2, and 3.4 +/- 0.1 mmol/L for control, insulin, and fasted animals, respectively. Blood beta-hydroxybutyrate concentrations were identical (0.5 +/- 0.1 mmol/L) for control and insulin-treated animals, but more than doubled in concentration in the fasted animals (p less than 0.001). Mortality rates during hypoxia-ischemia were higher in the insulin-treated animals (30%) than in either the fasted (4%) or control (0%) animals (p less than 0.05). Fasted animals showed a significant reduction in hypoxic-ischemic brain damage as compared with either the insulin-treated or control animals. Insulin-treated animals were not significantly different from controls. The findings indicate that 1) insulin induced hypoglycemia does not provide a protective effect on perinatal hypoxic-ischemic brain damage, as in adults; and 2) fasting adequate to produce hypoglycemia and ketonemia improved neuropathologic outcome.