The objective of this study was to compare the results of the two most commonly used techniques for near-field and specific absorption rate (SAR) evaluations. The comparison was performed by independently assessing the near-field of a commercially available DCS1800 cellular phone in two different laboratories. The measurements were conducted at the Swiss Federal Institute of Technology with the near-field scanner DASY2 equipped with specialized near-field E and H field probes, while the numerical evaluation was performed using the finite difference time domain technique at the Institute for Mobile and Satellite Communications. The close near-fields in free space, inside a flat phantom filled with tissue simulating liquid and inside complex and homogeneous head phantoms, were compared. The quantitative results of most test configurations tallied well. Although the current distribution of the simulation was only slightly different from that of the physical phone, differences of the spatial peak absorption were greater than > 2 dB in some positions of the phone with respect to the scatterer. This demonstrates the great sensitivity of the spatial peak SAR on the current distribution on the device both in terms of magnitude and geometrical distributions. For numerical compliance tests this places high requirements on the numerical phone model with respect to the accuracy of representing the near-field RF characteristics of the phone.