The aim of this study was to compare results obtained from two commercially available immunoassay kits for intact proinsulin. Fasting and post-prandial samples were obtained from both healthy subjects and patients with type 2 diabetes mellitus and assays were carried out according to the manufacturers' instructions. Coefficient of variation of the duplicates in both assays was acceptable with the MLT Intact Proinsulin assay giving slightly better overall precision. Regression analysis indicated a good correlation between the assays (r=0.97), however, a procedure better designed to compare analytical methods demonstrated a considerable lack of agreement for some samples. Dilution of samples in the Dako assay greatly affected the results when compared to samples assayed undiluted, whereas in the MLT assay, dilution of samples produced the expected results. Repeat comparison, assaying samples neat in the MLT assay and diluted 1:5 in the Dako assay, resulted in a considerable improvement in the agreement between the Dako and MLT assays. This study underlines the importance of the use of validation procedures which demonstrate quantitative analytical recoveries from a variety of specimens over the working range of the assay method in question.