A recently marketed passive hemagglutination (PHA) test (Rubacell, Abbott Laboratories) was compared to the hemagglutination inhibition (HI) test for the detection of antibody to rubella virus. The performance of the PHA system in determining immunity to rubella was evaluated by comparing the PHA results and HI results on 1, 086 randomly selected sera submitted for routine premarital and prenatal testing. Of the 1, 079 specimens assayable by both procedures, 1, 053 of the results (97.6%) were in agreement on initial testing. When the 26 initially discrepant specimens were retested for clarification, there was final agreement in 1, 067 of the specimens (98.9%). Twelve specimens were classified as persistently discrepant (five were not retested by PHA) and seven were unassayable by HI. Seven of the specimens with discrepant results were PHA positive and HI negative, and five were PHA negative and HI positive. Discussion of the two tests with respect to technical difficulty, cost, and controls is also included.