Human growth hormone (hGH) therapy has been used for the possible improvement of adult height in individuals with idiopathic short stature (ISS) for more than 20 years. However, given its heterogeneity, and the distinction between 'partial' or 'transient' GH deficiency, the exact definition of ISS is difficult. Since recombinant hGH became available, individuals with all types of GH deficiency have been extensively treated, often without the need for diagnostic justifications. On the other hand, for ISS individuals GH treatment has so far only been possible in a few countries, or in the context of clinical trials. If hGH is certainly effective in individuals with severe GH deficiency, its efficacy is lower and with a high individual-to-individual variability of response in those with non-severe GH deficiency and with ISS. The most important variables associated with a favorable growth response to hGH therapy in GH- and non-GH-deficient individuals are first-year growth response, younger age at start of treatment, difference at start from target height standard deviation score, GH dose, and other variables not necessarily associated with GH peak levels after provocative stimuli. A better choice of individuals to be treated successfully will improve the cost-effectiveness of the treatment, making it more acceptable to the scientific community and other stakeholders.