Wireless power transfer (WPT) via magnetic resonance offers efficient electrical power transfer, making it an increasingly attractive option for charging electric vehicles (EVs) without conventional plugs. However, EV charging requires a transfer power in order of kW or higher, resulting in a higher-leaked magnetic field than conventional wireless systems. The leaked magnetic field is nonuniform, and the assessment in terms of the limit prescribed in the guideline is highly conservative because it assumes that a person standing in free space is exposed to a uniform field. In such cases, an assessment should be performed using the limits of the internal electric field, as it is more relevant to the adverse health effects, whereas its evaluation is time-consuming. To mitigate this over-conservativeness, international product standards introduce a spatial averaging method for nonuniform exposure assessment. In this study, we investigate assessment methods, especially for measurement points of nonuniform magnetic field strength leaked from the WPT system. Various spatial averaging methods are correlated with the internal electric field derived from electromagnetic field analysis using an anatomically based human body model. Our computational results confirm a good correlation between the spatially averaged magnetic and internal electric fields. Additionally, these methods provide an appropriate compliance assessment with the exposure guidelines. This study advances our understanding of the suitability of spatial averaging methods for nonuniform exposure and contributes to the smooth assessment in WPT systems.