Rapid increase in the use of numerical techniques to predict current density or specific absorption rate (SAR) in sophisticated three dimensional anatomical computer models of man and animals has resulted in the need to understand how numerical solutions of the complex electrodynamics equations match with empirical measurements. This aspect is particularly important because different numerical codes and computer models are used in research settings as a guide in designing clinical devices, telecommunication systems, and safety standards. To ensure compliance with safety guidelines during equipment design, manufacturing and maintenance, realistic and accurate models could be used as a bridge between empirical data and actual exposure conditions. Before these tools are transitioned into the hands of health safety officers and system designers, their accuracy and limitations must be verified under a variety of exposure conditions using available analytical and empirical dosimetry techniques. In this paper, empirical validation of SAR values predicted by finite difference time domain (FDTD) numerical code on sphere and rat is presented. The results of this study show a good agreement between empirical and theoretical methods and, thus, offer a relatively high confidence in SAR predictions obtained from digital anatomical models based on the FDTD numerical code.