Evaluation of the immunity to rubella by the hemagglutination inhibition (HI) test and an enzyme-linked immunosorbent assay (ELISA), in a series of 1,697 sera collected from women of childbearing age, showed a 65% agreement; 33% were ELISA-positive and HI-negative at a 1:16 dilution and only 2% were ELISA-negative and HI-positive. Among 241 women given the rubella vaccine following an HI titer of less than 1:16, no change was observed in 208 in the postimmunization HI test, but only 11 were negative when ELISA was used. The retesting by ELISA of 110 preimmunization sera showed that 67 (61%) were positive, and therefore the vaccine had been unnecessarily administered. In a series of 72 HI tests that did not permit interpretation because of a nonspecific reaction, ELISA provided quite clear results. ELISA appears to be a sensitive technique for the evaluation of the true immune status to rubella in sera with low HI levels (less than 1:16), as well as in circumstances where the HI test reacts nonspecifically.