In studies on experimental infection of Rickettsia mooseri (=R. typhi) in Xenopsylla cheopis and laboratory rats, it was found that 10 days after the infectious feeding, the fleas were voiding feces that were infective to rats upon inoculation. The feces remained infective for at least the duration of the experiment, and a quantity as small as 0.2 µg of feces would result in seroconversion of 67% of the rats upon inoculation. Fleas were capable of transmitting the infection to rats as early as seven days after feeding on rickettsemic rats, but the rate of transmission was much higher late in the course of rickettsial development in the flea, e.g., virtually 100% by day 17. Fleas transmitted R. mooseri infection even when they fed on the host for a maximum of 30 min and were removed from the rats at least 25 min before they could be expected to deposit any feces. These and other data suggest that R. mooseri may be transmitted by X. cheopis by the feeding process, and not merely through contact with infective feces or crushed fleas. The ambient temperature had a profound effect upon rickettsial growth in the fleas. At 18°C, the rickettsial content of the fleas was below detectable levels for at least ten days and remained low throughout, whereas at 24°C and 30°C the rickettsial titer was consistently two or three times greater. However, if, after six days, the fleas were transferred from an environment of 18°C to one at 24°C or 30°C, the rickettsial growth increased by two or three logs within one week.