In the present study of twenty-two patients with Cushing's syndrome, serum insulin-like growth factor (IGF)-I concentrations were normal to high with an increased mean IGF-I concentration, 40% above that of healthy subjects of the same age (p<0.001). Serum IGF-II concentrations were normal. The morning serum IGF binding protein (IGFBP)-l concentrations were within the range of healthy controls. IGFBP-1 was inversely correlated to the IGF-I concentration (p<0.001) and to the 24 h urinary cortisol excretion (p <0.005) with a combined R squared value of 0.58. In contrast to IGFBP-l, serum IGFBP-2 and IGFBP-3 concentrations were elevated by 1.89±1.78 sd and 0.92±0.78 sd (mean±2 sd), respectively. Although increased, the serum IGFBP-2 concentration was inversely correlated to the IGF-I concentration (r= −0.67, p<0.001). Immunoreactive IGFBP-3 was increased in proportion to IGF-I and the molar ratio [IGFBP-3]: [IGF-I]±[IGF-II] was close to unity (1.04±0.14), as that of healthy subjects. In serum from patients with Cushing's syndrome, with increased immunoreactive IGFBP-3, there was a corresponding increase in intact glycosylated 40–43 kDa IGFBP-3 as determined by Western ligand blotting. Neutral size chromatography of serum from patients with Cushing's syndrome showed that IGF-I and IGFBP-3 immunoreactivity were predominantly found at the elution volume of the ternary 1 50 kDa IGF-I/IGFBP-3/acid labile subunit complex and a similar pattern was displayed by normal serum. We conclude that the catabolism and impairment of growth in patients with Cushing's syndrome cannot be attributed to the suppression of the GH-endocrine IGF-I axis. Furthermore, our findings suggest that the previously reported increased IGF inhibitory activity, of low molecular weight, in Cushing's serum could not be attributed to IGFBP-3, IGFBP-3 fragments or IGFBP-1. The possible role of IGFBP-2 or other IGFBPs in blocking the effects of IGFs has to be further evaluated.