Numerous physiological factors modulate GH secretion, but these variables are not independent of one another. We studied 40 younger (20-29 yr.; 21 men and 19 women) and 62 older (57-80 yr.; 35 men and 27 women) adults to determine the contributions of several demographic and physiological factors to the variability in integrated 24-h GH concentrations. Serum GH was measured every 10 min for 24 h in an enhanced sensitivity chemiluminescence assay. The predictor variables included: age group (young or old), gender, abdominal visceral fat (by computed tomography), total body fat mass and percentage body fat by dual-energy x-ray absorptiometry, serum IGF-I, fasting serum insulin, 24-h mean estradiol and testosterone, and peak oxygen uptake by graded exercise (treadmill) testing. Multiple ordinary least squares regression analysis was used to quantitatively assess the individual contribution that each predictive measure made to explain the variability among values of integrated 24-h GH concentrations while in the presence of the remaining predictors. The model explained 65% of the variance in integrated 24-h GH concentrations. Abdominal visceral fat (P < 0.002) and fasting insulin (P < 0.008) were consistently important predictors of integrated 24-h GH concentrations independent of age group, gender, and all other predictor variables. Although serum IGF-I was an important overall predictor of integrated 24-h GH concentrations (P = 0.002), this relationship was present only in the young subjects and was modulated by gender. The remaining variables failed to contribute significantly to the model. We conclude that abdominal visceral fat and fasting insulin are important predictors of integrated 24-h GH concentrations in healthy adults, independent of age and gender. Serum IGF-I is an important predictor of integrated 24-h GH concentrations in young but not older subjects. Bidirectional feedback between each of these three factors and GH secretion may account for the strong relationships observed.