Serum levels of total insulin-like growth factor I (IGF-I) correlate with growth hormone (GH) secretory status and are a useful parameter in the diagnostic evaluation of GH deficiency. Serum total IGF-I levels represent the combined quantity of free or unbound IGF-I and IGF-I that is bound to specific IGF binding proteins. Free IGF-I (fIGF-I), which is postulated to be the bioactive fraction, accounts for only a small fraction of the total amount. We have recently developed a new immunoradiometric assay (IRMA) for plasma fIGF-I and have investigated fIGF-I in relation to GH status. The simple, non-extraction assay procedure involves the capture of unbound IGF-I by anti-IGF-I antibody coated to polystyrene beads and detection by a radiolabelled anti-IGF-I antibody directed to a separate epitope. Preliminary studies demonstrated that the fIGF-I IRMA does not measure IGF-I that is complexed to IGF-binding proteins and that the equilibrium between the free and bound fractions is not disturbed during the assay. Free IGF-I levels were compared to total IGF-I levels measured in the same IRMA after acid-ethanol extraction of the samples. Normal levels of fIGF-I from infancy through adulthood were found to have a close correlation with total IGF-I levels, with the lowest levels occurring in infancy and peak levels during puberty. Patients with complete GH deficiency had low levels of both fIGF-I and total IGF-I, with 94% and 100% of the levels below the 5ht percentile for age, respectively. On the other hand, approximately 90% of patients with normal IGF binding protein-3 levels among partial GH deficiency and normal short children had free and total IGF-I levels above the 5th percentile for age. These data indicate that the clinical utility of plasma fIGF-I measurements is similar to measurements of total IGF-I in the evaluation of childhood GH deficiency.