Insulin is produced by beta cells in pancreatic islets of Langherans via a complex process of proteolytic conversion. A precursor molecule, proinsulin, is transported to the Golgi apparatus where it is packed into secretory granules. Maturation of the secretory granules is associated with conversion of proinsulin to insulin and C-peptide by enzymatic cleavage. Secretion of insulin into the bloodstream is accompanied by the release of small amounts of proinsulins. Insulin immunoassays consist of radioimmunoassays using polyclonal antisera which cross-react with proinsulins, and two-site assays using monoclonal antibodies. These immunometric assays have led to improvements in specificity and sensitivity as compared to radioimmunoassays. To determine reference values and limits, insulinaemia must be measured in normoglycaemic subjects with a normal body weight. Moreover, as insulinaemia is most often measured during stimulation tests, reference values must also be determined for the most common tests such as the oral glucose tolerance test or the intravenous glucose tolerance test. We report the analytical characteristics of insulin assays and review reference values and their interpretation. Wide-scale use of insulin assays remains a subject of research rather than a diagnostic application. Spontaneous hypoglycaemia, a disorder which can be caused by hyperinsulinism, insulinoma, insulin autoimmune syndrome and non-insulin-mediated factors, is almost the only clinical indication for the measurement of plasma insulin. Diabetes is diagnosed solely on the basis of chronic hyperglycaemia. Thus, measurement of plasma insulin has no clinical value in the diagnosis or management of diabetic patients, with the exception of rare cases including the syndrome of severe insulin resistance and abnormalities in beta-cell secretory products. Otherwise, insulin measurement is used in experimental investigations to study the pathophysiology of various disorders, especially diabetes. The reference and range of plasma insulin values are not yet clearly established, and the range of concentrations reported in the literature remains unsatisfactory. There is a need to standardise results and thereby improve comparability among studies.