Roads are important features in the landscape that can result in biodiversity loss. Roadkill surveys are a common method used to determine the vulnerability of species to roads. Surveys are done by patrolling roads in vehicles or walking, and collecting data associated with any wildlife mortality. Like any method, the validity of data collected via roadkill surveys hinges on avoiding or accounting for major biases. Carcass persistence is considered the most important factor that can bias the accuracy of roadkill estimates, and small mammals are particularly prone to bias due to their size. The purpose of this study was to test if season (winter vs. spring), land use (agricultural fields vs. forests), and placement on road (shoulder vs. middle of road) affect the detectability of small mammal roadkill. I used a variety of roads in southeastern Wisconsin, USA, that differed in the aforementioned factors to drop dead feeder rats (purchased commercially) to simulate roadkill of small mammals. I conducted surveys every 12 hours to check the detectability of the feeder rats. My results show that 90% of carcasses disappeared within 24 hours. In addition, season, land use, and position all affected the removal rates of small mammal roadkill. Anecdotal evidence suggests that many carcasses were removed by crows (Corvus brachyrhynchos). My results support the notion that the mortality of small mammals may be underestimated in roadkill surveys, and precautions should be taken to minimize or account for these biases.