The aim of this study was to assess body proportions in children with short stature born small for gestational age (SGA) before and during 6 years of growth hormone (GH) treatment. A prospective randomised double-blind dose-response study comparing the effects of 3 vs. 6 IU GH/m2/day. Seventy-nine children with short stature (height SD-score < -1.88) born small for gestational age (birth length SD-score < -1.88). Before and during GH treatment, height, sitting height (SH), hand (Hand) and foot length (Foot), biacromial (Biac) and biiliacal diameter (Biil) were measured. All results were adjusted for age and sex, and expressed as SD-scores (SDS) using reference values for healthy Dutch children. To describe the size of SH, Hand, Foot, Biac, and Biil in relation to height, these values were adjusted for the SDS of height. At baseline, these short children had small hands and feet and narrow shoulders and pelvis compared to healthy peers. Height and SH, were, however, even more affected. Consequently, on average, these children had relatively large hands and feet, and relatively broad shoulders and pelvis compared to their height, but a normal sitting height in proportion to height. In most of the individuals, the values for body proportions were, however, within the normal range. During 6 years of GH treatment the SD-scores of all measurements increased significantly towards values more close to zero. The mean size of Hand, Foot, and Biil decreased in proportion to height. The mean SH increased relatively more than height, however, to values well within the normal range. The mean Biac in relation to height had not changed after 6 years of GH treatment. No differences in the 6-year changes in body proportions were found between the two GH dosage groups. Untreated short children born small for gestational age have, on average, relatively large hands and feet, and broad shoulders and pelvis, but a normal sitting height compared to height. The increase in height during 6 years of GH treatment is accompanied by an improvement of the proportions of the size of hands, feet, and biiliacal diameter, in relation to height. The increase in height appeared to be the result of the increase in sitting height as well as leg length, but the sitting height SD-score increased slightly more than that of leg length. The changes in body proportion during GH treatment were dose-independent. Thus, 6-year continuous GH treatment with either 3 or 6 IU/m2/day in children with short stature born small for gestational age does not negatively influence body proportions.