Railroads use train speed measurements to assess operational efficiency and safety. The recent availability of low-cost GPS receivers presents an opportunity for massive cost reduction in monitoring continuously the speed and position of equipment across the entire network. GPS receivers estimate speed from geospatial position updates. However, low-cost GPS receivers can produce relatively large errors in position updates, thereby producing similar errors in speed estimates. Studies tend to focus on characterizing GPS receiver errors in urban road settings. Subsequently, railroads know very little about the nature of GPS errors along rural train routes. Smartphones nowadays have all the necessary sensor capabilities needed to test and validate a low-cost speed monitoring system. This study characterizes speed errors by using multiple smartphones onboard a hi-rail vehicle. The authors describe the data collected, the data processing algorithm developed to estimate speed, and the error quantification by comparing speed estimates to vehicle speedometer measurements reported by the hi-rail vehicle operators.