We have previously shown that the major correlates of growth following growth hormone (GH) therapy in growth hormone-deficient (GHD) children are changes in circulating insulin-like growth factor-I (IGF-I) and IGF binding protein-3 (IGFBP-3), suggesting a synergistic interaction between IGF-I and IGFBP-3 (1). The first aim of this project was to examine the molecular forms of IGFBP-3 and the acid-labile subunit (ALS), and to assess the changes in these molecular forms during GH administration to GHD children. Plasma samples from prepubertal GHD patients, prior to therapy and during the first year of GH treatment, were subjected to Western ligand and immunoblot analysis. Densitometric analysis of Western ligand blotting (WLB) showed a 76% increase in IGFBP-3 (p = 0.02), but a 56% decrease in 36-kDa IGFBP-2 (p = 0.03) during GH therapy. Western immunoblot (WIB) analysis of IGFBP-3 revealed the presence of intact (40- to 45-kDa doublet) as well as a proteolyzed (28-kDa) form of IGFBP-3 in the serum of GHD and healthy children. Both immunoreactive forms of IGFBP-3 increased by 64% during GH therapy (intact p = 0.003; proteolyzed p = 0.0001). WIB analysis of the ALS showed an 84-to 86-kDa doublet, which increased by 41% with GH therapy (p = 0.01). The response to GH therapy, as measured by the height velocity standard deviation score (SDS) adjusted for bone age, correlated with the percent change in total IGFBP-3 (r = 0.772, p = 0.002 by WIB), intact IGFBP-3 (r = 0.845, p = 0.0005 by WLB; r = 0.541, p = 0.05 by WIB), and proteolyzed IGFBP-3 (r = 0.703, p = 0.007), as well as with the percent change in ALS (r = 0.813, p = 0.014). The second aim of this project was to assess the changes in distribution of the immunoreactive forms of IGFBP-3 and IGF-I among the ternary (ALS/IGFBP-3/IGF) complex, the binary (IGFBP-3/IGF) complex, and uncomplexed IGF during the first year of GH therapy, and to explore further the correlation with growth response to GH. Plasma samples, prior to therapy and after the first year of GH treatment, were separated by neutral size-exclusion chromatography and then subjected to IGFBP-3 immunoradiometric assay (IRMA), IGFBP-3 WIB, and IGF-I IRMA analysis. IGFBP-3 increased in both the ternary (p < 0.0001) and binary (p = 0.01) complexes, but there was a shift in the percentage of IGFBP-3 from the binary to the ternary complex during GH therapy. Both intact and proteolyzed forms of IGFBP-3 were found in both the ternary and binary complexes, but the shift occurred primarily for the proteolyzed (28-kDa) form (p = 0.001). There was a significant increase in IGF-I in the ternary (p = 0.001) and binary (p = 0.005) complexes, but not in uncomplexed IGF-I. The percentage of IGF-I in the ternary complex increased (p = 0.006), whereas the percentage of uncomplexed IGF-I decreased (p = 0.02), during GH therapy. Growth rate, assessed by the height velocity SDS for bone age, correlated best with the changes in ternary complex IGFBP-3 (r = 0.72, p = 0.01) and ternary complex IGF-I (r = 0.56, p = 0.10). In conclusion, GH treatment of GHD children results in significant increases of intact, proteolyzed, and total IGFBP-3, as well as an increase in ALS, which all correlate with the growth response to GH therapy. In addition, GH treatment results in increases in ternary complex IGFBP-3 and IGF-I, which also correlate with the response to therapy. We suggest that the formation of the ternary complex may be a determining factor in the somatic growth response.