To detect suspension faults and the wheel-rail condition, a fault-detection-and-isolation system for a railway vehicle bogie -- with a model-based estimation method using a Kalman-Bucy filter and isolation scheme -- was developed. The advantage of this system is that many faults can be detected by using a small number of sensors. Time-history simulations were carried out to investigate the performance of the new fault-detection and isolation system. The simulation results show that faults in the suspensions and changes in the wheel-rail interface can be detected from the residual given by the Kalman-Bucy filter. They also show that each fault can be identified by the fault-isolation scheme. It is thus concluded that the model-based estimation method using a Kalman-Bucy filter and isolation scheme provides an effective basis for a fault-detection-and-isolation system for a railway vehicle bogie.