Underwater sensor networks (USN) are used for tough oceanographic missions where human operation is dangerous or impossible. In the common mobile USN architecture, sensor nodes freely float several meters below the surface and move with the force of currents. One of the significant challenges of the mobile USN is localization. In this paper, we compare the performance of three localization techniques; Dive and Rise Localization (DNRL), Proxy Localization (PL) and Large-Scale Localization (LSL). DNRL. PL and LSL are distributed, range-based localization schemes and they are suitable for large-scale, three dimensional, mobile USNs. Our simulations show that, DNRL and LSL can localize more than 90% of the underwater nodes with high accuracy while LSL has higher energy consumption and higher overhead than DNRL. The localization success and accuracy of PL is lower than the other techniques however it can localize underwater nodes faster when small number of beacons are employed. (C) 2010 Elsevier B.V. All rights reserved.