We performed glucagon stimulation tests in 59 normally growing siblings of children who died from sudden infant death syndrome. These investigations were performed to exclude a possible metabolic disorder (found in 4 siblings) as an underlying cause of sudden infant death syndrome. The remaining 55 siblings (32 boys and 23 girls) provide control data for this age range. Testing was performed at 0800 h after a 15-h fast. The median age was 98 days (range, 13-349 days). Plasma glucose and serum cortisol, insulin, and GH were determined before and 30, 60, 90, 120, 150, and 180 min after im injection of 0.1 mg/kg glucagon. No side-effects were observed during the procedure. Asymptomatic hypoglycemia was noted in 11% of the infants at least once between 120-180 min. Basal and peak GH concentrations were greater than 10 micrograms/L in 31% and 80% of the infants, respectively. There was a significant negative correlation between age and basal GH concentration [Spearman's rank correlation coefficient (rs) = -0.37; P < 0.01]. There was a significant correlation between age and glucagon-stimulated cortisol at 120, 150, and 180 min (rs) = 0.41; P < 0.005), but not between age and changes in glucose levels. There was no significant correlation between age and basal cortisol or peak GH concentrations and no difference between boys and girls for any of the variables studied. In conclusion, the glucagon stimulation test is well tolerated in very young subjects. The peak GH response to glucagon injection is independent of age between 0.5-12 months. The age-related increase in the glucagon-stimulated cortisol response despite a similar decrease in glucose suggests the existence of a postnatal maturation in the response of the pituitary-adrenal axis to stress.