Fifty-eight plasma specimens from 30 patients who had undergone presurgical radioimmunoscintigraphy with 111In-labeled anti-carcinoembryonic antigen (CEA) murine monoclonal antibody (Mab) and who had no clinical evidence of disease after surgical resection showed increased concentrations of CEA (greater than or equal to 5 micrograms/L) in plasma when studied with the previously available commercial CEA enzyme immunoassay (EIA) from Roche. The possible role of anti-murine antibody (HAMA) interference was addressed by adding mouse IgG (mIgG) to the plasma (2 g/L) before assay. Fifteen specimens (26%) showed no change in CEA (reflecting a true increase as shown by the original results), 22 (38%) showed a decrease in CEA of greater than 15% but remained positive (reflecting an artefactual increase), and 21 (36%) became CEA-negative (less than or equal to 5 micrograms/L; reflecting a false increase). Subsequently, we assayed the same samples with a modified version of this CEA EIA kit and 47 specimens remained CEA positive (greater than 5 micrograms/L): 25 (53%) were truly increased, 12 (26%) remained artefactually increased, and 10 (21%) continued to show a false increase. The degree of interference in the original EIA kit correlated with the plasma concentration of HAMA (P less than 0.005). All artefactually and falsely increased CEA values observed in both kits were corrected by addition of polyclonal mIgG or of a mixture of IgG1, IgG2a, and IgG2b Mabs before assay. This correction is important in the follow-up of patients who receive murine Mabs for treatment or diagnosis.