PURPOSE: Growth hormone (GH) treatment has been used to improve growth in short children who were born small for gestational age (SGA). The aim of this study was to investigate the long-term efficacy of GH treatment in these children. METHODS: Data from a multicenter observational clinical trial (ClinicalTrials.gov NCT01604395, LG growth study) were analyzed for growth outcome and prediction model in response to GH treatment. One hundred fifty-two children born SGA were included. RESULTS: The mean age of patients born SGA was 7.13 +/- 2.59 years. Height standard deviation score (SDS) in patients born SGA increased from -2.55 +/- 0.49 before starting treatment to -1.13 +/- 0.76 after 3 years of GH treatment. Of the 152 patients with SGA, 48 who remained prepubertal during treatment used model development. The equation describing the predicted height velocity during 1st year of GH treatment is as follows: the predictive height velocity (cm) = 10.95 + [1.12 x Height SDS at initial treatment (score)] + [0.03 x GH dose (ug/kg/day)] + [0.30 x TH SDS at initial treatment (score)] + [0.05 x age (year)] + [0.15 x Weight SDS at initial treatment (score)] +/- 1.51 cm. CONCLUSIONS: GH treatment improved growth outcome in short children born SGA. We also developed a prediction model that is potentially useful in determining the optimal growth outcome for each child born SGA. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT01604395.