Fleas are small wingless hematophagous insect that are frequently infesting on rodents and other small mammals while acting as vectors of many zoonotic diseases including plague, a disease challenging public health in many part of the world including Tanzania. 291 rodents from nine species were captured with Sherman traps in farm land, peridomestic areas, bush and forest buffer zone across wet and dry season in plague and non-plague foci villages. Captured rodents were anaesthetized and 190 fleas comprising four species were collected and morphologically identified with available dichotomous key. Dinopsyllus lypusus were (46.32%) , Ctenophthalmus spp (26.84%), Xenopsylla brasiliensis (16.32%) and Xenopsylla cheopis (10.53%). 38.42% of fleas were collected from Mastomy natalensis, 22.63% from Lemniscomys striatus and 18.42% from Rattus rattus. Highest flea infestation prevalence was found on R.rattus and was strongly associated with X.cheopis and X.brasiliensis. Specific flea index of X.cheopis on R.rattus was (01) in plague foci and (<0.5) in non-plague foci. Result of GLM final model indicated that flea abundance was significant influenced by rodent species ( p < .001), season ( p= .031), habitats ( p= .02), rodent weight ( p < .001), rodent sex ( p < .001) and plague locations ( p= .02). There was significance difference in variation of flea abundance between rodent sexes (W = 9158.5, p = .009) and weak positive correlation between rodent’s weight and fleas abundance ( R= 0.17, p< 0.05). Despite that, specific flea index of X.cheopis on rats in both plague foci and non-plague foci villages were not indicating alarming condition that would require urgent control of fleas, still society should consistently adhere to rodent and fleas control methods especially in farm land and peridomestic areas where flea abundance is high and human activities are customarily implemented.