Dissociation between GH bioactivity (bio-GH) and GH immunoactivity (immuno-GH) is due to the heterogeneity of the molecule: the measurements do not always provide reliable information on the bio-GH. We studied the ratio of bio-GH and immuno-GH during pharmacological secretion tests in 211 sera to study the concentration-response curve of the assay (C1), 16 samples of normally growing subjects with idiopathic short stature (C2), 13 samples from patients with GH deficiency (GHD1) and 6 samples of 3 patients with GHD and normal provocative tests (GHD2). GH bioactivity was determined by the Nb2 cell proliferation assay (bio-GH) and immuno-GH by a time-resolved immunofluorometric assay (IFMA) (immuno-GH). A non-linear negative relationship between the serum bio-GH/immuno-GH ratio and serum immuno-GH was observed in C1. In log-log plotting representation, two cut-off lines were drawn: a vertical cut-off line separating above-below cut-off serum peak immuno-GH values in provocative tests, and a diagonal cut-off line separating normal-abnormal serum bio-GH/immunoGH ratio; four areas were defined. GHD1 had normal ratios, but below cut-off peak immuno-GH responses. P2 and P3 of Group GHD2 had abnormal ratios in samples with low serum immuno-GH but only P2 had autosomal dominant mutation. P1 had the same autosomal dominant isolated GHD as P2 but a low normal ratio. Our data underline the importance of relatively low serum GH concentrations in mediating GH biological actions. An abnormal serum bio-GH/immuno-GH ratio might explain certain cases of GHD and might be useful in detecting abnormal circulating isoforms of GH in patients with growth failure.