OBJECTIVE Anti-tumor necrosis factor (TNF) therapy is known to decrease disease activity of juvenile idiopathic arthritis (JIA), but its effect on longitudinal growth in relation to puberty is not clear. We studied longitudinal growth in response to etanercept treatment in prepubertal and pubertal patients with JIA. METHODS Out of 52 children treated with etanercept, we studied 20 prepubertal and 11 early/midpubertal patients adherent to treatment for at least 1 year. We collected data on growth and glucocorticoid medication and calculated each patient's height standard deviation score (SDS) in relation to the mid-parental height, the change of this value (DeltahSDS) from 1 to 0 and 0 to 1 year of treatment, and the change between the DeltahSDS values to assess growth improvement. RESULTS In the prepubertal group, the relative height SDS (mean +/- standard error of the mean) was 1.8 +/- 0.2, 2.1 +/- 0.3, and 1.9 +/- 0.3, and in the pubertal group 1.1 +/- 0.4, 1.3 +/- 0.3, and 1.1 +/- 0.3 at 1, 0, and +1 year of treatment, respectively. The DeltahSDS before etanercept was 0.3 +/- 0.1 in prepubertal and 0.2 +/- 0.2 in pubertal patients. Over the first year with etanercept, DeltahSDS was +0.2 +/- 0.1 in prepubertal (p = 0.001 vs before etanercept; paired Student t-test) and +0.2 +/- 0.1 in pubertal patients (p = 0.071). Nevertheless, most prepubertal (17/20) and pubertal (8/11) patients had improved growth (DeltahSDS) in response to etanercept treatment when analyzed individually. The need for intraarticular glucocorticoid injections was negatively correlated to the improved growth (p = 0.001). CONCLUSIONS TNF inhibition with etanercept improved growth in a majority of patients with JIA. Our data demonstrate that growth improvement with etanercept was independent of the pubertal growth spurt.