Growth failure is common during long term treatment with glucocorticoids (GC) due to blunting of GH release, insulin-like growth factor I (IGF-I) bioactivity, and collagen synthesis. These effects could theoretically be reversed with GH therapy. The National Cooperative Growth Study database (n = 22,005) was searched for children meeting the following criteria: 1) pharmacological treatment with GC and GH for more than 12 months, 2) known type and dose of GC, and 3) height measurements for more than 12 months. A total of 83 patients were identified. Monitoring of glucose, insulin, IGF-I, IGF-binding protein-3, type 1 procollagen, osteocalcin, and glycosylated hemoglobin levels was performed in a subset of patients. Stimulated endogenous GH levels were less than 10 microg/L in 51% of patients and less than 7 microg/L in 37% of patients. The mean GC dose, expressed as prednisone equivalents, was 0.5 +/- 0.6 mg/kg day. Baseline evaluation revealed extreme short stature (mean height SD score = -3.7 +/- 1.2), delayed skeletal maturation (mean delay, 3.1 yr), and slowed growth rates (mean, 3.0 +/- 2.5 cm/yr). After 12 months of GH therapy (mean dose, 0.29 mg/kg x weeks), mean growth rate increased to 6.3 +/- 2.6 cm/yr, and height SD score improved by 0.21 +/- 0.4 (P < 0.01). During the second year of GH therapy (n = 44), the mean growth rate was 6.3 +/- 2.0 cm/yr. Prednisone equivalent dose and growth response to GH therapy were negatively correlated (r = -0.264; P < 0.05). Plasma concentrations of IGF-I, IGF-binding protein-3, procollagen, osteocalcin, and glycosylated hemoglobin increased with GH therapy, whereas glucose and insulin levels did not change. The following conclusions were reached. The growth-suppressing effects of GC are counterbalanced by GH therapy; the mean response is a doubling of baseline growth rate. Responsiveness to GH is negatively correlated with GC dose. Glycosylated hemoglobin levels increased slightly, but glucose and insulin levels were not altered by GH therapy.