The normal values of insulin-like growth factors (IGFs) after extraction, their binding proteins, and the high affinity GH-binding protein are not well established in infancy or childhood. We report the relationship between serum IGF-I, IGF-II, their binding proteins IGFBP-1 and IGFBP-3, and GH-binding protein in 600 normal Spanish children who were divided into 5 groups according to Tanner stage: I, 150 males and 102 females; II, 40 males and 42 females; III, 45 males and 45 females; IV, 42 males and 55 females; and V, 23 males and 56 females. Serum IGF-I levels increase slowly during childhood in both sexes, exhibiting a dramatic increase during puberty and a significant decline [P < 0.001, by analysis of variance (ANOVA)] during adulthood. The pubertal peak occurs approximately 2 yr earlier in girls than in boys. In contrast, serum IGF-II levels remain stable throughout childhood, showing no pubertal peak. In boys, there is a significant decline in IGF-II levels during adulthood (P < 0.001). Serum IGFBP-3 levels show a pattern similar to that of IGF-I, with a significant increase during childhood and a significant decline during adulthood (P < 0.001, ANOVA) in both males and females. In contrast, serum IGFBP-1 levels decrease dramatically during childhood in both boys and girls (P < 0.001 and P < 0.005, respectively, by ANOVA). A significant decline in serum GH-binding protein levels is observed between prepubertal and pubertal children of both sexes (P < 0.001). There is a close linear correlation between the sum of serum IGF-I plus IGF-II levels vs. serum IGFBP-3 (r = 0.724; P < 0.0001). In contrast, there is a nonlinear correlation between serum IGF-I vs. serum IGFBP-3 (concave curve) as well as between serum IGF-II and serum IGFBP-3 (convex curve). A negative correlation was found between serum IGF-I vs. IGFBP-1 (r = -0.51; P < 0.0001) as well as between the sum of serum IGF-I plus IGF-II vs. IGFBP-1 (r = -0.47; P < 0.0001), but not between serum IGF-II and IGFBP-1. These data emphasize that when these tests are performed in the clinic, their interpretation should be based upon age- and sex-specific criteria.