The study of the cardiolipin antigen demonstrates the relationships between the three constitutents—two phospholipids, cardiolipin and lecithin, with cholesterol, essentially chemically pure substances—which are necessary to its standardization for the complement-fixation test. The inhibitive effect of cardiolipin upon the hemolytic activity of complement increased progressively with the amount of cardiolipin tested. Inhibition was relatively slight with smaller amounts of cardiolipin which in mixture with lecithin and cholesterol possessed marked antigenic activity in the presence of syphilitic serum. When strongly reacting sera were tested with cardiolipin alone, some inhibition of complement was observed, although this was of relatively slight degree in relation to the total reactivity of the serum when tested with a suitably adjusted mixture of cardiolipin, lecithin, and cholesterol. In the presence of normal or nonreacting serum or of suitable amounts of lecithin, cardiolipin did not inhibit the hemolytic activity of complement. Lecithin alone was not anticomplementary. Neither did lecithin inhibit complement in the presence of nonreacting or reacting serum with the exception of one specimen, the reaction with which was relatively slight in degree as compared with that obtained with the same serum and a suitably standardized antigen. Lecithin and cardiolipin, when mixed in suitable proportions and dispersed by methods commonly used with noncholesterolized antigen, were highly active antigens. The proportion that provided maximal reactivity was approximately five parts of lecithin to one of cardiolipin. The total concentration of the mixtures in alcoholic solution did not affect appreciably the reactivity of the saline dilutions; the optimal reaction was dependent in each instance on the total amount of cardiolipin and lecithin present in the saline dilution used. When cholesterol was added to mixtures of cardiolipin and lecithin, a definite sensitization resulted as indicated by the fact that under these conditions reactions of like degree were detected by much smaller quantities of cardiolipin and lecithin. When lecithin and cardiolipin were used in a ratio of 5:1, maximal sensitization by cholesterol was approximated by a ratio of cholesterol:lecithin of 3.4:1. The antigenic activity of such mixtures, like that of mixtures without cholesterol, was, in these studies, independent of the absolute concentration of antigen in alcohol; the saline dilutions that proved optimal for use in tests were approximately proportional to the concentration of antigen in alcohol, making the absolute amount of antigen used in the test about the same. When an amount of lecithin was used in excess of the ratio 5:1, the ratio of cholesterol providing maximal reaction was not altered; when the amount of lecithin was considerably diminished, however, the ratio cholesterol:lecithin required for maximal reactivity progressively increased and was accompanied by a gradual increase in anticomplementary reaction, the degree of which exceeded that exhibited by similar quantities of cardiolipin in the absence of cholesterol. It has thus been possible by titrating the serologic reactivity of known syphilitic serum with cardiolipin antigen that contains three essentially chemically pure substances to determine the ratios between these three components that will give the maximal reaction under the conditions of the test. A ratio of lecithin:cardiolipin of 5:1 and of cholesterol:lecithin of 3.4:1 appears to be a conservative adjustment for practical use. A ratio of lecithin:cardiolipin of about 1.5:1 and of cholesterol:lecithin of 5 or 6:1 appears to approximate the upper limits of sensitization in the experiments of this study.