Rabbit sperms were injected intravenously into female rabbits in single doses of 5 million per animal, and in weekly or semi-weekly doses of 65-288 million sperms per injection. The sera of these animals were tested weekly for 3 months. Similarly, both stock and Wistar Institute female rats were given 100,000, 200,000 and 300,000 rat sperms intramuscularly at 4-day intervals. After 10 days, serum tests were made and further injections of 3-15 million sperms were given at 3-4-day intervals. Female guinea pigs were given 5 semi-weekly intraperitoneal injections amounting to a total of 283 millions of sperms each. Serum tests were made for 6 weeks. In each experiment, the animals serum was tested for isospermotoxin before the antigen (macerated epididymides and testes in 0.9% NaCl solution) was administered. As indicated by the motility of the sperms in a hanging drop of serum, the rabbit sera were entirely negative, while the sera of the treated rats were indistinguishable from those of the controls; both stopped sperm motility in less time than did the saline. The sera of every guinea pig stopped motility in the sperms much before the normal animals serum or saline. The toxic effect, which is specific for guinea pig sperms, was abolished by heating to 56[degree] C. for 30 min., but was restored by adding suitable complement[long dash]normal guinea pig serum.