Abstract: Radio frequency (RF) heating has been expected as an alternative to traditional thermal treatments for postharvest pasteurization and disinfestation, due to its rapid heating, high thermal efficiency, volumetric heating， and selective heating. Corn is one of the most primary food crops in terms of both large yield and wide use in the world. Since a large amount of postharvest loss was caused by the invasion of mold and insects every year, especially in developing countries, it is highly urgent to develop RF heating for corn kernels. However, it needs to accurately predict the temperature when formulating parameters for the process of RF heating, where the corn temperature is the primary determinant of product quality. In this study, a three-dimensional geometric model of corn kernels was constructed using the simplified structure of germ and endosperm, and an RF heating mathematical model for corn kernels was developed using the heterogeneity of thermos-physical and dielectric properties in the COMSOL multi-physics software. A validated model was used to simulate the distribution of temperature and electric field, thereby exploring the changes of RF selective heating in corn kernels with different postures and moisture content. The results showed that the simulated and experimental temperatures were highly consistent, with a maximum relative error of 3.47%, indicating that the model could well predict the behavior of RF selective heating in corn kernels. The cold spots appeared in the center of the rectangular container, while the hot spots occurred in its corners and edges during RF heating, showing non-uniform heating. Specifically, the temperature difference between the corn kernels at the corner and center gradually increased as time proceeded. In surface temperature, the temperature of the germ was higher than that of endosperm, indicating the preferential heating of the germ. The geometric effect of the upright corn kernel was the largest, followed by the oblique and side-standing ones, where the preferential heating of germ was the greatest, as the enhanced intensity of electric field and power density of germ, whereas, that of the flat corn kernel was the smallest. Moreover, the temperature difference between germ and endosperm in corn kernels with the moisture content of 13.0% increased gradually during RF heating, while that in corn kernels with the moisture content of 16.5% and 20.0% increased rapidly first and then decreased. When the corn kernels were heated to 55 ℃, the preferential heating of germ in corn kernels with the moisture content of 16.5% was the largest, followed by those with the moisture content of 13.0%, and those with 20.0% moisture content was the smallest. Therefore, the posture and moisture content of corn kernels can be used to adjust RF heating and reduce the damage of RF selective heating on the germ quality. The current work can contribute to understanding the RF heating behavior in corn kernels, thereby serving as a useful guidance for the application of RF heating in low-moisture corn kernels.