To assess possible side-effects of GH treatment with supraphysiological doses on carbohydrate (CH) metabolism in girls with Turner syndrome (TS) during long term GH treatment and after discontinuation of GH treatment, the results of oral glucose tolerance tests and hemoglobin A1c measurements were analyzed in 68 girls with TS participating in a randomized dose-response trial. These previously untreated girls, aged 2-11 yr, were randomly assigned to 1 of 3 GH dosage groups: group A, 4 IU/m2 x day (-0.045 mg/kg x day); group B, first year ,4 IU/m2 day; thereafter, 6 IU/m2 x day (approximately 0.0675 mg/kg x day); group C, first year, 4 IU/m2 x day; second year, 6 IU/m2 x day; thereafter, 8 IU/m2 x day (approximately 0.090 mg/kg x day). After the first 4 yr, girls 12 yr of age or older started with 5 microg/kg BW-day 17beta-estradiol for induction of puberty. To assess the effects of long term high dose GH treatment on CH metabolism, the 7-yr data from the oral glucose tolerance tests in 9 girls of group C were evaluated (group C1). To determine whether the changes in CH metabolism during GH treatment would persist after discontinuation of GH treatment, the data for 28 girls who had reached adult height (group A, n = 9; group B, n = 10; group C, n = 9) were evaluated at baseline, after 4 yr of GH treatment, and 6 months after discontinuation of GH. Seven-year data for group C1 showed that glucose levels did not significantly change during GH treatment, whereas fasting insulin levels as well as glucose-induced insulin levels increased significantly. The data for the 28 girls who were treated with GH for a mean (SD) period of 85.3 (13.3) months demonstrated that the GH-induced higher insulin levels decreased to values close to or equal to pretreatment values after discontinuation of GH treatment. Changes in CH variables were not significantly related to the GH dose. Hemoglobin A1c levels never showed an abnormal value. The prevalence of impaired glucose tolerance was low, and none of the girls developed diabetes mellitus. In conclusion, long term GH treatment with dosages up to 8 IU/m2 x day in girls with TS has no adverse effects on glucose levels, but induced higher levels of insulin, indicating relative insulin resistance. The increased insulin levels during long term GH treatment decreased after discontinuation of GH treatment to values close to or equal to pretreatment values. Although the reversibility of the effects of long term GH is reassuring, the consequence of long term hyperinsulinism is still unknown.