An IgG antibody avidity assay which uses urea to modify a commercial enzyme-linked immunosorbent assay (ELISA) has been investigated for its ability to distinguish primary human cytomegalovirus (CMV) from recurrent or long-term infection. Twenty-four immunosuppressed solid organ transplant patients were studied. The avidity indices for IgG to CMV were low for 12 out of 13 patients with primary infection (mean 18%), high for all 11 patients with long-term infection (mean 85%), and the 1 patient with primary infection showing an intermediate avidity index (51%) was found to have acquired passively large amounts of CMV immunoglobulin, presumably of high avidity, during therapy. From the results, low and high avidity indices were defined as lying between 0-34% and 60-100%, respectively, and it was thus clear that the avidity assay can discriminate between primary and recurrent or long-term CMV infection. The avidity indices of eight of the immunosuppressed organ transplant patients with primary infection were followed in serial serum samples over time and IgG antibody to CMV was found to take at least a year to mature to high avidity in contrast to the 2-6 months expected for normal subjects. This finding provides evidence that immunosuppression has subtle, hitherto unsuspected, effects on humoral immunity to CMV in addition to the well-known depression of cell-mediated responses. It is concluded that this reliable avidity assay will be of importance in the diagnosis of CMV infection and in elucidating the pathogenesis of CMV-induced disease in organ transplant recipients.