The cytomegalovirus (CMV) antibody status of 79 cardiac transplant recipients and their respective donors was determined by a latex-agglutination technique and by a CMV IgG ELISA technique on the same samples. Although good agreement between the two methods was found in recipient samples the latex method produced significantly more positive results in donor samples than did the ELISA test. Clinical and serological follow-up of the patients concerned over the first 6 months after transplantation suggested that the ELISA results were correctly predictive of donor transmitted primary CMV infection and that, therefore, the latex test was producing false positive readings. Further investigation is needed to discover the reason for these discrepancies. The use of donor samples taken at the time of organ retrieval may have prejudiced the latex test results. Until these problems are resolved it is suggested that the latex agglutination technique should not be the sole method adopted to assess donor CMV antibody status.