The aim of this study was to investigate if contrast agents interfere with the performance of an immunoradiometric assay (IRMA) in vitro for serum tumour-associated antigen. Each of five carcinoembryonic antigen (CEA)-positive sera, CA-130-positive sera and tissue polypeptide antigen (TPA)-positive sera was mixed with six contrast agents: Ioversol 350, Iopamidol 370, Iomeprol 300, Iomeprol 400, Iohexol 300 and Gadopenteic acid in 50:50, 50:20, 50:5.0, 50:1.0, 50:0.5 and 50:0.1 microl proportions. Following IRMA, the interference of binding rates in each mixture was calculated, and the serum concentrations of CEA, CA-130 and TPA were estimated and compared with the originals. All contrast agents used were able to inhibit the binding rate with IRMA and the inhibition rates were in proportion to the amount of contrast agent. The detection of serum concentrations of CEA, CA-130 and TPA was significantly inhibited in the mixtures with more than 5.0 microl of contrast agent in all cases. Apart from Iomeprol 400, there was no significant inhibition of detection at the lowest concentrations of contrast agents. Iomeprol 400 was the strongest inhibitor and Gadopenteic acid the weakest inhibitor for each IRMA of the contrast agents employed. In conclusion, our results demonstrate that contrast agents may reduce the immunoreaction of antibody and antigen and lead to in vitro inhibition during immunoassays. It would be unwise to perform any plasma/serum immunoassay on a sample collected within 24 h of the administration of contrast agent considering the pharmacokinetics.