[Objective] As clayey soil contains much clay and clay loam, it is rather complex in dielectric polarization. However, not much study has been done on mixed dielectric models for measurement of water content.[Method] In this study, based on the theory for bilinear dielectric measurement, dielectric spectra were determined, separately of soil samples, 0, 5%, 10%, 15%, 20%, 25%, and 30% in volumetric water content, prepared out of each of the 4 clay soils different in texture were determined, separately, within the frequency band of 0.001~3 GHz.[Result] Dielectric value (the real part of complex dielectric constant and apparent dielectric constant) is relatively stable within 300.4~2 952 MHz, but varies sharply with volumetric water content. The band of 1 050~1 503 MHz is the ideal frequency band in the frequency domain for predicting clay water content, and 1503 MHz is the optimal frequency point. At this point, two empirical models for predicting water contents in clayey soils in the frequency domain were constructed. In analyzing clay samples and validation samples, comparisons of the values predicted with the real part of the complex dielectric model with the measured values of water content show that the predictions are all higher than 0.9600 in R², lower than 0.0190 in RMSE, and higher than 5.000 inn RPD. In analyzing clay soil samples, comparisons between the values predicted with the apparent dielectric model formula and the measured values show that R²=0.9669, RMSE=0.0176, and RPD=5.515. Compared with the Topp model, Roth model and Malicki model, the apparent dielectric model yields values much closer to the measured value. In analyzing verification samples, comparison between the values predicted with the apparent dielectric model and the measured values show that R²=0.9537, RMSE=0.0208, and RPD=4.602.[Conclusion] The bilinear theory is quite high in applicability to the measurement of clay soil moisture content with the dielectric method. The two empirical models constructed in this paper are fairly high in accuracy in predicting water content in soils clayey in texture.