Objective: Treatment of GH-deficient (GHD) children with higher doses of recombinant human GH (rhGH) than conventional ones has been reported to result in higher growth velocity and increased final height. These findings, however, were observed by comparing large but heterogeneous groups of children. We wanted to verify whether the same results could be obtained in two groups of appropriately well-matched children with isolated GHD treated with high vs conventional doses of rhGH. Methods: Out of two cohorts of GHD children, cohort A (on a weekly rhGH dose of 0.3 mg/kg body weight) and cohort B (on a weekly rhGH dose of 0.15 mg/kg body weight), we selected two groups, each including 13 patients, who before treatment were matched for age, sex and height standard deviation score (SDS). They were followed up until final height. Results: Final height SDS was significantly higher in group A ð20:45^0:36 (S.D.) vs 21:07^0:7; P ¼ 0:008Þ; as well as height gain SDS ð1:81^0:58 vs 1:23^0:62; P ¼ 0:002Þ: The difference between final height SDS and target height SDS was positive only in group A and significantly higher in group A than in group B ð0:33^0:51 vs 20:46^0:7; P ¼ 0:01Þ: Glucose tolerance was always normal in the group treated with higher doses. Conclusion: The final height of children treated with higher doses of rhGH is increased, also in relation to their genetic target. The economical burden of this choice of treatment, however, has to be taken into account when evaluating the results.