The level and distribution of radiofrequency energy absorbed in a child's head during the use of a mobile phone compared to those in an adult head has been a controversial issue in recent years. It has been suggested that existing methods that are used to determine specific absorption rate (SAR) and assess compliance with exposure standards using an adult head model may not adequately account for potentially higher levels of exposure in children due to their smaller head size. The present study incorporates FDTD computations of locally averaged SAR in two different anatomically correct adult and child head models using the IEEE standard (Std. C95.3-2002) SAR averaging algorithm. The child head models were obtained by linear scaling of the adult head model to replicate the conditions of previous studies reported in the literature and also by transforming the different adult head models based on data on the external shapes of children's heads. The tissue properties of the adult and corresponding child head models were kept the same. In addition, modeling and experimental measurements were made using three spheres filled with a tissue-equivalent mixture to approximate heads of increasing size. Results show that the peak local average SAR over 1 g and 10 g of tissue and the electromagnetic energy penetration depths are about the same in all of the head models under the same exposure conditions. When making interlaboratory comparisons, the model and the SAR averaging algorithm used must be standardized to minimize controversy.