High degree of variability has been reported in the measurement of growth hormone (GH) using immunoassays (IA) with different sensitivity and specificity. The aim of this study was to evaluate the maximal GH level (GH max) in response to provocative test by measuring GH using three different IA and to compare their usefullness in the diagnosis of GH deficiency (GHD). Twenty nine short children, 22 without clinical apparence of GH deficiency(SS) and 7 GHD, were studied Sequential arginine-clonidine test were performed in two separate occasions under either estradiol (E2 1-2 mg/d, p.o. for three days) or placebo pretreatment. GH max was determuned in 58 serum samples by RIA (in house assay, minimal detection dose, MDD: 1 μg/L). IRMA (Serono, MAIA, MDD: 0.15 μg/L) and IFMA (DELFIA, MDD: 0.01 μg/L).Results: Significant correlations were found among the different methods used (IRMA vs RIA, r=0.97; IFMA vs RIA, r=0.93; IFMA vs IRMA, r=0.97 p<0.001). Estrogen pretreatment have a similar stimulatory effect on both mean levels of GH max and the lower limit of 95% confident limits measured by different IA. On placebo: 11.1 = 10.4 (3.1-40.5) 10.4 ± 12.6 (2.2-49) y 4.4 ± 11 (0.71-27) μg/L; on E2: 18.0 ± 10.7 (7.2-45), 17.8± 12.4 (6.3-50.1) and 8.2 ± 7.3 (2.3-28.5) μg/L (p<0.05) using RIA. IRMA and IFMA respectively. Without E2-priming, 3 to 4 out of 7 GHD patients showed values within the 95% CL for the SS group using the three different IA. Estradiol-primed GH max reached the normal values in only 1/7 by RIA and 0/7 by both IRMA and IFMA. Conclusions: Estradiol-priming enhanced GHD diagnosis by GH provocative test using three different IA. However, there was not a significant improvement using more sensitive and specific assays.