Focusing on the localization problem of indoor mobile robots, a credibility based adaptive federated filter localization method is designed. Firstly, considering the features of mobile robots and the indoor environment, a localization system with multiple sensors is designed, and the sensors are configured accordingly. To handle the information from multiple sensors, a federated filter is introduced, and each sub-filter is designed. On the basis of it, data fusion is made to the measured data for obtaining the pose estimation of mobile robots. Then, considering the characteristics of a particular sensor, the credibility factors of its measured data are given by using fuzzy reasoning. Meanwhile, based on the credibility factors, an adaptive fusing algorithm of the federated filter is designed, and the localization result of indoor mobile robots is given. Finally, simulation results are given to verify the effectiveness of the designed localization algorithm.