The single radial hemolysis (SRH) test was compared with the hemagglutination inhibition (HI) test for establishing rubella immune status and diagnosing recent infection. Correlation between mean SRH diameters and HI titers greater than or equal to 1:8 was high (R = 0.99). It is suggested that a level of greater than or equal to 5 IU represents low-level antibody and that greater than or equal to 15 IU is a conservative threshold for designation of immunity. Of 343 sera tested, only 1 false-positive was found by SRH with the 5 IU cutoff level. The SRH test could detect serum antibody levels as low as 2.5 IU, whereas 15 IU was generally the limit of resolution of the HI test. Data from sucrose density gradient fractionation of serum demonstrated that neither rubella-specific immunoglobulin M (IgM) nor early postinfection HI-reactive IgG was detected by SRH. However, diagnostic changes in antibody titer measured by SRH were in general greater than those measured by HI. The SRH test showed diagnostic titer changes in some sera containing specific IgM for which no such changes were detected by the HI test. A 2.5-mm difference in hemolytic zone diameters (a fourfold rise in international units) between acute- and convalescent-phase serum pairs was chosen as being of diagnostic significance. This difference was less than the minimum observed difference of 2.9 mm from 59 serum pairs showing diagnostic changes by HI and far exceeded (greater than 3.6 standard deviations) the within-test and individual variability seen for 95 pregnant women from whom samples were obtained during each trimester.