Due to increased availability of growth hormone (GH) for the treatment of short stature, its use has been proposed for a number of conditions besides classic GH deficiency. We have studied growth response during a one year treatment period with 14 IU/m2/week of GH in a heterogenous group of 24 short children with various conditions associated with short stature (SDS for body height ranging between -2.2 and -4.4). Thirteen children could be classified as "responders" with growth rate increments of 2 cm/yr or more above pretreatment growth rates, and 11 children as "non-responders". The children were measured regularly both by stadiometry and knemometry at weekly intervals. GH stimulation by insulin, arginine and spontaneous overnight secretion of GH, and SM-C generation were evaluated in the children and found to be of no predictive value for the individual responsiveness to GH administration except in one boy with classic GH deficiency. However, serial measurements of the lower leg length provided useful information for individual predictions in 21 out of the 24 children as early as 10 weeks after the start of the GH treatment.