A capture enzyme-linked immunosorbent assay (ELISA) for detection of virus-specific immunoglobulin M (IgM) antibody was developed which used a panel of labeled monoclonal antibodies to rubella virus hemagglutinin. The rapidity of the test system was increased by using, after 1-h incubation of the test serum, a second 1-h incubation of the serum with a mixture of viral antigen and labeled monoclonal antibody. The new assay was tested for specificity on 371 human sera from people without any recent contact with rubella virus; of these, 66 were sera selected from people with rheumatoid factor or IgM antibody to human cytomegalovirus, Epstein-Barr virus, or other viruses. In parallel, the new assay was performed on 191 sera from patients having recent contact with rubella virus. Results were compared with those obtained by an indirect ELISA method on IgM serum fractions, using purified rubella virus as a solid phase. Of the 371 sera tested for specificity, 5 (1.3%) gave false-positive results with indirect ELISA (1 rheumatoid factor, 2 heterophil antibody, and 2 human cytomegalovirus sera positive for IgM), and none were false-positive with the capture assay. Two sera from a patient with primary cytomegalovirus infection, which were positive for rubella IgM antibody with both methods and were initially interpreted as false-positive, were finally considered to be true-positive, since they were reactive only in the presence of IgM antibody and viral antigen. Of the 191 sera from 92 patients (84 patients with acute rubella, four newborns from mothers with rubella during pregnancy, and four vaccinees), 136 (71.2%) were found to be positive for IgM by direct ELISA, and 128 (67.0%) were positive by capture ELISA; 12 sera drawn during the first 2 days of disease, or at least 40 days after onset (or after vaccination), were detected only by indirect ELISA, and 4 sera were detected only by capture ELISA. Thus, specificity and sensitivity, respectively, were 100 and 91.4% for capture ELISA and 98.6 and 97.1% for indirect ELISA. However, when the number of patients was considered, 86 were detected as IgM positive by indirect ELISA, and 87 were detected positive by capture ELISA. The overall agreement between the two assays was 96.2%. Capture ELISA using monoclonal antibody appears preferable over indirect ELISA on IgM serum fractions because of its higher specificity and shorter time for test performance; furthermore, there is no need for serum fractionation or virus purification for the capture ELISA.