Accurate localization of interventional devices, for example, needles and brachytherapy seeds, is desired for interventional procedures. MRI is usually considered unsuitable for this purpose, as the induced signal voids and signal pile-ups do not necessarily represent the exact location of the devices. Center-out radial sampling with off-resonance reception (co-RASOR) has been shown to solve this problem by repositioning the signal pile-up into the geometrical center of the interventional devices. However, the multiple acquisitions required for co-RASOR resulted in a low efficiency and unsuitability for near real-time interventional purposes. Herein, we aim to increase the efficiency of co-RASOR by relying on multiple off-resonance reconstructions of a single acquisition rather than on multiple acquisitions. The soundness of this approach is shown by demonstrating the equivalence of acquisition co-RASOR and reconstruction co-RASOR, both theoretically and experimentally. An algorithm is proposed and evaluated to obtain the geometric centers of the devices, while suppressing the background. This procedure is shown to be effective, in vitro as well as ex vivo, and to yield signal intensity increases in the order of 150–400% of the average signal, in the geometric center of a brachytherapy seed and a needle, respectively. The geometric accuracy of the resultant images is confirmed by computed tomography. Magn Reson Med, 2013. © 2012 Wiley Periodicals, Inc.