Morbidly obese subjects are characterized by multiple endocrine abnormalities and these are paralleled by unfavorable changes in body composition. In obese individuals, either 24-h spontaneous or stimulated GH secretion is impaired without an organic pituitary disease and the severity of the secretory defect is proportional to the degree of obesity. The GHRH+arginine (GHRH+ARG) test is likely to be the overall test of choice in clinical practice to differentiate GH deficiency (GHD) patients. Similarly to other provocative tests, GHRH+ARG is influenced by obesity per se. Therefore, a new cut-off limit of peak GH response of 4.2 microg/l in obese subjects has been recently assumed. The aim of the present study was to investigate the reciprocal influence between decreased GH secretion and body composition in a group of 110 morbidly obese subjects, using the new cut-off limit of peak GH response to GHRH+ARG test for these subjects. In our study, GHD was identified in 27.3% of the obese subjects, without gender difference. In GDH obese subjects body mass index (BMI), waist circumference, waist-to-hip ratio (WHR), fat mass (FM), and resistance (R) were higher while reactance (Xc), phase angle, body cell mass (BCM), IGF-I, or IGF-I z-scores were lower than in normal responders (p<0.001). In all obese subjects, GH peak levels showed a negative correlation with age, BMI, waist circumference and FM, and a positive correlation with IGF-I. In the stepwise multiple linear regression, waist circumference and FM were the major determinants of GH peak levels and IGF-I. In conclusion, using the new cut-off limit of peak GH response to GHRH+ARG test for obese subjects, about 1/3 morbidly obese subjects were GHD. GHD subjects showed a significantly different body composition compared with normal responders, and the secretory defect was correlated to different anthropometric variables with waist circumference and FM as the major determinants.