Human growth hormone (GH) mediates longitudinal bone growth and also exerts a variety of other biological effects, e.g. lactogenic, insulin-like, diabetogenic, lipolytic, protein-anabolic and sodium/water-retaining effects. Investigation of the structure-function relationship of human GH has been attempted by epitope mapping using monoclonal antibodies and by systematic point mutations of the human GH molecule. The diagnosis of GH-related disorders is complicated by the fact that different commercial immunoassay kits give widely differing results. This phenomenon cannot be explained by lack of standardization, but has to be attributed to the epitope specificities of the antibodies employed in the assay techniques and the spectra of different molecular forms of human GH recognized by them. The information required by a clinician in ordering a GH determination is the bioactivity of all the human GH forms in the sample, rather than the immunoreactivity with a given set of antibodies. Therefore, the aim of future GH immunoassays must be the identification of antibodies that bind with high affinity only those forms of human GH that bind and activate the GH receptor. Epitope mapping by monoclonal antibodies appears to be the ideal approach to this goal. It has recently been shown that one molecule of human GH binds two molecules of GH receptor, and dimerization of the receptor by its ligand is a prerequisite for biological function. Improvements in understanding the structure-function relationship of the human GH molecule and of the interplay of the hormone with its receptor make it conceivable that recombinant analogues of human GH will be designed to inhibit the effects of GH excess in acromegaly. Furthermore, analogues of human GH could be designed to exhibit a modified spectrum of biological activities compared with wild-type human GH.