Abstract: The "high impedance" at a substation may become a main cause of the failure of the fault location in the distribution network. A distributed fault location method is proposed for distribution networks under measuring devices' optimal configuration for resolving the problem. The key features of the proposed method are that under the globally optimal placement of the measuring devices, the distribution fault section location is realized based on the temporal characteristics of the faulty transient signals. Firstly, the improved l_{2, 1} principal component analysis is used to identify the sensitive nodes in the distribution network, which will be taken as the globally optimal configuration locations for the measuring devices. Then the fault monitoring domain is constructed and the fault monitoring augmented matrix is obtained according to the configuration results. Furthermore, the fault section identification rules are presented to identify the fault section, and the fault point is located according to the basic theory of the traveling wave method. An improved IEEE 34-node feeder system is modeled with the PSCAD/EMTDC. The simulation results show the proposed method can accurately identify the fault sections and has strong robustness to the transition impedance. The locating results are not affected by the fault types and the fault positions.