The aim of the present study was to evaluate retrospectively the influence of various auxological and laboratory parameters on final height in a group of GH-deficient children after replacement therapy and to compare their final height with that of a group of short children with normal GH secretion and hence not treated. The final height was evaluated of 83 patients (51 males and 32 females) affected by idiopathic isolated GH deficiency and treated with recombinant human GH (hGH) for 2-7 years. Inclusion criteria at the start of treatment were short stature (mean height for chronological age in standard deviation score (SDS) -2.21) due to idiopathic isolated GH deficiency (GH peak < 8 micrograms/l after two pharmacological tests and/or mean GH concentration < 3.3 micrograms/l during the night) and treatment with recombinant hGH for at least 2 years at a dose of 15-20 U/m2 per week by s.c. injection for 6 or 7 days/ week. Mean chronological age at diagnosis was 12.2 +/- 1.7 years; 35 were prepubertal and 48 pubertal. The final height of 51 untreated short stature (mean height for chronological age in SDS -2.13 at diagnosis) subjects (42 males and 9 females: 29 prepubertal and 22 pubertal at diagnosis with mean chronological age 11.6 +/- 2.4 years) with normal GH secretion was also evaluated. In the treated subjects final height SDS was higher than that of the untreated group (-1.3 vs -1.7 SDS; P = 0.01). Both treated and untreated subjects showed a final height lower than target height, but 39% of the treated subjects vs only 20% of the untreated group (P = 0.035) had a final height greater than target height. In the treated subjects this percentage was higher in the patients improving their height for bone age in the first years of therapy. While treated females showed a positive correlation only between target and final height (P = 0.0001), in treated males final height correlated with the Bayley-Pinneau prediction at diagnosis, height for chronological age and bone age at diagnosis and target height. Patients who started therapy before puberty also showed these correlations with data calculated at the onset of puberty, together with a correlation with chronological age at the onset of puberty. When considering the influence of GH response at tests on final height, the percentage of subjects exceeding target height increased progressively according to the severity of the GH deficiency. There was no difference in height gain between the patients starting therapy before or during puberty. The height gain, however modest, obtained by our treated patients, the number of patients with final height greater than target height and the favourable comparison with the untreated short-stature subjects represent a promising result, which could be improved by personalizing treatment.