Clinical and basic studies have documented a high susceptibility to pneumococcal infection in asplenic humans and animals. It has been suggested that autotransplantation of splenic tissue might be a method of providing host resistance when total splenectomy is necessary. However, the effect of splenic autograft has remained controversial. This study was performed to evaluate the most effective site and amount of splenic autograft using rats. Rats were divided into five groups for the purpose of determining the site of splenic autotransplantation: splenectomy, sham operation, implantation into the omental pouch, intraperitoneal implantation, and intramuscular implantation. For determining the amount for autotransplantation, the rats were divided into seven groups: splenectomy, sham operation, and implantations of 25, 50, 100, 200, or 300 mg of splenic tissue. All animals were challenged with Streptococcus pneumoniae type 6, 16 weeks after surgery. Howell-Jolly bodies appeared postsplenectomy, but disappeared in the implanted rats 16 weeks after the operation. Histologically, the implanted tissue was indistinguishable from that of a normal spleen. Pneumococcal clearance from the bloodstream and survival rate were significantly higher in rats implanted in the omental pouch as compared with splenectomized rats. Intraperitoneal and intramuscular implanted rats did not show a significant difference from the splenectomized rats. More than 50% of splenic tissue for autograft showed a significant increase in pneumococcal clearance and survival rate as compared with that of splenectomized rats. It was suggested that the most effective site of autotransplantation is the omental pouch and approximately 50% of the whole spleen would be necessary for prevention from sepsis.