Abstract: A RSS-based location method based on weighted multidimensional scaling is proposed in this paper. Based on the relation between received signal strength (RSS) and the distance from emitted source to the receiver, the distance information can be first obtained. Next, the source location can be modeled as multidimensional scaling and the multidimensional scaling matrix can be decomposed by subspace method. According to the orthogonality of signal subspace and noise subspace, linear equations for location parameter can then be obtained. By using the statistical characteristic of the errors for the linear equations, the weighted least square solution for source location can be finally found. Computer simulation shows the efficiency and practicality for this method. The method can attain the CRB at moderate SNR.