To compare interpretations of growth hormone (GH) provocative tests in laboratories using six different GH immunoassays (one enzymeimmunometric assay (EIMA, assay 1), one immunoradiometric assay (IRMA, assay 5), one time-resolved fluorimmunometric assay (TRFIA, assay 3) and three radioimmunoassays (RIAs, assays 2, 4 and 6)), aliquots of peak samples from GH provocative tests were distributed between the four participating laboratories, quantified in the respective immunoassays and interpreted according to the cut-off values for provocative tests defined for each assay method. There was a high degree of relative correlation between the different assays, but absolute GH estimates differed. Assays 2, 4, 5 and 6 yielded almost identical GH levels. Assays 1 and 3 yielded serum GH levels approximately 39% and 70%, respectively, of those of assays 2, 4, 5 and 6. Although the absolute GH level measured in the various assay methods varied, there was good agreement between the interpretation of a given sample among the participating laboratories. This indicates that the differences in estimates of serum GH concentration by different immunoassay systems can be compensated for when cut-off values are defined for each method.