Because of the data lost and gross error of ultra wideband (UWB) positioning system in a complex environment, the tracking accuracy of SINS/UWB integrated positioning system is seriously decreased. Aimed at above problem, this paper proposes an SINS/UWB integrated positioning method based on fault tolerant detection for UWB. Firstly, the stationary state detection model is built with measurement of SINS using the Median filter. Moreover, the measured bias of SINS can be calibrated with Least Square algorithm for the stationary state. Secondly, the fault tolerant judgement model is used to detect the UWB, which includes the measured data lost and gross position error caused by None Line of Sight and interference signal. The fault tolerant integrated positioning system can be achieved by Kalman Filter using detected positioning results of UWB. Finally, some experiments are conducted to evaluate the proposed method using experimental platform of SINS/UWB integrated positioning system. The experimental results show the fault tolerant integrated positioning system can effectively detect the data lost and gross error for UWB, meanwhile, the fault tolerant positioning system can guarentee high tracking accuracy for a duration time of 6s for UWB inaccurate state.