Abstract: Multidimensional scaling (MDS) is an efficient data analysis technique. Applying the MDS technology to sensor localization is a novel idea. An NMDS-RSSI (nonmetric MDS and received signal strength indication) localization algorithm is presented, which runs nonmetric MDS on RSSI between nodes instead of distance to localization, so it avoids the process of transforming RSSI into distance by other localization approaches based on RSSI, and can reduce errors caused by this kind of transformation. By using nonmetric MDS, the localization approach provides robust localization even in the presence of random RSS fluctuations due to multi-path fading and shadowing through extensive simulations and a representative set of real experiments.