The variables which affect the qualitative and quantitative aspects of the interaction between fowl erythrocytes and rubella virus were defined and evaluated. The sensitivity of the hemagglutination reaction is at least tenfold greater at pH 6.0 to 6.5 than it is at pH 7.2, the pH of dextrose-gelatin-Veronal which is widely used as diluent for rubella hemagglutination (HA) and hemagglutination inhibition (HAI) tests. In the presence of 0.001 M Ca++ ion the reaction is two to eight times more sensitive than it is in the absence of Ca++ ion. The development of agglutinated and non-agglutinated RBC patterns that are stable and clearly distinguishable from one another is critically dependent upon the concentrations of gelatin and/or albumin in the diluent. The diluents employed in previously published HAI test procedures differ widely with respect to pH, Ca++, gelatin and/or albumin concentration. The lack of comparability of HAI test results among laboratories may in part reflect these differences in diluent composition. In an effort to correct such difficulties, optimal conditions for the agglutination of fowl RBC by rubella virus were defined. A 0.14 M NaCl solution buffered with N-2-hydroxyethylpiperazine-N'-2'-ethanesulfonic acid (HEPES) at pH 6.05 to 6.15, containing 0.001 M CaCl2, 1.0% bovine serum albumin and 0.00025% gelatin was shown to provide conditions for a maximally sensitive, stable and clear test. Under these conditions the outcome of the hemagglutination test is solely a function of the concentration of hemagglutinin in the test specimen.