In murine Schistosoma mansoni infections, fewer adult worms develop in male than in female mice infected with the same number of cercariae. To evaluate a potential role for testosterone in this phenomenon, testosterone levels were manipulated in groups of CBA/J mice that were then infected and monitored for survival rates, worm burdens, organomegaly, and egg production. By 16 weeks of infection, more than 80% of mice in groups with low levels of testosterone (untreated females, castrated males, or carrier-treated castrates) were dead, while less than 40% of those in groups with high levels of testosterone (sham-castrated males, testosterone-treated castrates, or testosterone-treated female mice) succumbed to infection. The mean number of worms recovered from mice in the low testosterone level groups was comparable among groups, and significantly greater than that from those in high-testosterone-level groups. The degree of organomegaly observed correlated strongly with worm burden, but the number of hepatic eggs per female worm did not differ significantly between groups. When male mice were castrated or sham-castrated 5 weeks after S. mansoni infection, no significant differences in host survival occurred. Furthermore, female mice treated with testosterone demonstrated reduced worm burdens if the testosterone was given 10 days prior to infection but not if the testosterone was given 10 days or 5 weeks after infection. Thus, the host sex bias observed in parallel-infected male and female mice appears to be related to the presence of male gonadal tissue or testosterone early in infection, during the development of immature schistosomules.