The GH response to provocative stimuli in obese is often as low as in panhypopituitaric patients with severe GHD; however, IGF-I levels are normal or slightly reduced. In 53 patients with simple obesity (11 M and 42 F, age: 40.3+/-1.6 yr, BMI: 39.1+/-1.0 Kg/m2), we evaluated the GH response to GHRH (1 microg/kg iv)+arginine (ARG, 0.5 g/kg iv), and total IGF-I levels. The mean (+/-SE) GH peak after GHRH+ARG was markedly lower (74% reduction, p<0.0001) in obese (16.8+/-2.0 microg/l) than in normal subjects (62.7+/-4.3 microg/l). IGF-I levels in obese patients (134.0+/-7.6 microg/l) were lower (33% reduction, p<0.001) than in normal subjects (200.8+/-5.7 microg/l). Taking into account the 3rd centile limit of normal response, the GH response to GHRH+ARG was reduced in 62.3% (33/53) of the obese patients, and 21.2% (7/33) of them had low IGF-I levels. Assuming the 1st centile limit, it was reduced in 33.9% (18/53) obese subjects, and 22% (4/18) of them had low IGF-I levels. Considering 3.0 microg/L as arbitrary cut-off, the GH response was reduced in 5.7% (3/53) of the obese patients, and still one of them had low IGF-I levels. Our findings: a) confirm that the secretory capacity of somatotroph cells is often deeply impaired in obesity; b) demonstrate that IGF-I assay generally rules out severe impairment of GH/IGF-I axis in obese patients with marked reduction of the GH secretion; c) indicate that the percentage of obese patients with concomitant reduction of GH secretion and IGF-I levels is not negligible. Thus, IGF-I assay should be routinely performed in obese patients; those presenting with low IGF-I levels should undergo further evaluation of their hypothalamo-pituitary function and morphology, particularly in the presence of empty sella.