The prevalence of infectivity within a vector population is a critical factor in arthropod-borne disease epidemiology but it is difficult to estimate. In the case of bubonic plague, infective flea vectors contain large numbers of Yersinia pestis within a bacterial mass that blocks the flea's foregut, and only such blocked fleas are important for biologic transmission. A bacterial quantitation method could therefore be used to assess the prevalence of plague-infective (blocked) fleas in a population. We developed a standard, curve-based, competitive polymerase chain reaction (PCR) procedure to quantitate Y. pestis in individual fleas. The quantitative PCR (Q-PCR) method equaled a colony count reference method in accuracy and precision when evaluated using mock samples and laboratory-infected fleas. The Q-PCR was more reliable than colony count, however, for field-collected fleas and for blocked fleas collected after their death. In a sample of fleas collected from a prairie dog colony in the aftermath of a plague epizootic, 48% were infected but less than 2% contained numbers of Y. pestis indicative of blockage. The method provides a means to monitor plague epizootics and associated risks of flea-borne transmission to humans, and is applicable to the study of other vector-borne diseases.