The product of the obese (ob) gene, leptin, is an adipocyte-derived hormone that is involved in the regulation of appetite and body weight. This study was undertaken in order to describe the basal serum levels of leptin in prepubertal short children born small for gestational age (SGA) and their relationship with growth parameters, before and during growth hormone (GH) treatment. Eighty-nine prepubertal short children (66 boys, 23 girls; height standard deviation score (SDS), -5.4 to -2.0; age, 2.0 to 12.8 years) born SGA, 12 of whom (9 boys, 3 girls) had signs of Silver-Russell syndrome, were included in the study. Serum leptin concentrations were measured by radioimmunoassay. Leptin levels in the children born SGA were compared with those in a reference group of 109 prepubertal healthy children born at an appropriate size for gestational age (AGA). The mean (S.D.) change in height SDS was 0.11 (0.22) during the year before the start of GH therapy (0.1 IU/kg/day) and increased to 0.82 (0.44) during the first year (P < 0.001) and to 1.28 (0.59) during the 2-year period of GH therapy (P < 0.001). The children born SGA were significantly leaner than the reference group. An inverse correlation was found between leptin and chronological age in the SGA group (r = -0.31, P < 0.01). The mean serum level of leptin in the children born SGA who were older than 5.5 years of age was 2.8 micrograms/l which was significantly lower than the mean value of 3.7 micrograms/l found in the children born AGA of the same age range. The difference remained after adjustment of leptin levels for sex, age, body mass index (BMI) and weight-for-height SDS (WHSDSSDS). Leptin correlated with WHSDSSDS (r = 0.32, P < 0.001) and BMI (r = 0.36, P < 0.01) in the reference population, but not in the SGA group. No correlation was found between leptin and spontaneous 24-h GH secretion, insulin-like growth factor (IGF)-I or IGF-binding protein-3 levels, or with fasting insulin or cortisol levels. Leptin levels at the start of GH treatment were correlated with the growth response over both 1 year (r = 0.46, P < 0.001) and 2 years (r = 0.51, P < 0.001) of GH therapy. Using multiple regression analysis, models including leptin levels at the start of GH therapy could explain 51% of the variance in the growth response after 1 year and 44% after 2 years of GH treatment. In conclusion, serum leptin levels are reduced in short children born SGA and are inversely correlated with chronological age. Leptin concentrations correlate with the growth response to GH treatment and might be used as a marker for predicting the growth response to GH treatment.