An antigen prepared from toxoplasma-infected chorioallantoic membranes, improved by centrifugation at 13,000 r.p.m. for one hour which removed a nonspecific component capable of fixing complement with certain normal sera, proved to be the preparation of choice for toxoplasmic CF tests. Though it was possible to prepare an antigen of similar potency from mouse brain, the yield from one egg is at least 10 times more than from one mouse. The peritoneal exudate of mice freed of Toxoplasma and cells contains eight times as much antigen per unit volume as a 20% extract of mouse brain or chorioallantoic membranes, but its anticomplementary properties, which cannot be removed by high speed centrifugation, can be eliminated only by extensive dilution. The highly centrifuged and diluted peritoneal exudate, however, still contains a component which fixes complement with certain normal human sera. Sixteen children, aged 5 weeks to 11 years, with clinically recognized congenital toxoplasmosis and 17 mothers, with inapparent or clinically unrecognized infection except for having given birth to children with proved toxoplasmosis, were investigated for the presence and persistence of CF antibodies. With the improved antigen and the standardized method that was used, even titers of 1:2 and 1:4 were significant and invariably associated with toxoplasmic neutralizing antibody. The sera of all 17 mothers gave positive CF tests with titers ranging from 1:4 to 1:64, the higher titers predominating during the first two years after the delivery of the infected child. The sera of 14 of the 16 children (87.5%) with clinical evidence of congenital toxoplasmosis, confirmed by neutralization tests, gave positive CF tests with titers ranging from 1:2 to 1:128, the higher titers again predominating during the first two years after birth. One of the negative results, in a 5 week old child with active infection who died three days after the serum was obtained, was associated with a high (1:4,096) dye test titer for neutralizing antibody in its own serum, and a similar dye test titer (1:4,096) and a CF titer of 1:32 in the serum of its mother. The other negative result, in a 7 year old child with a dye test titer of 1:16, is believed to represent an instance of disappearance of CF antibody after an interval of seven years. Not one serum obtained from 24 children with congenital ocular or neurologic disturbances or both, not due to toxoplasmosis, gave a positive CF test with the toxoplasmic antigen; however, among 20 of their mothers, there were four sera with titers of 1:2, 1:2, 1:8 and 1:16. Toxoplasmic CF antibody can persist for at least six years, and in some instances even longer, in individuals with clinically recognized as well as inapparent or clinically unrecognized infections. Since at least seven of the mothers, who gave birth to children with congenital toxoplasmosis, subsequently gave birth to normal children at a time when the CF antibody titers in some of them were still high (32, 32, 32, 16, 8, 4, 4), it is clear that this antibody can persist in individuals in whom the infection has been eradicated or suppressed sufficiently to prevent its congenital transmission.