In order to study drying characteristics of contact ultrasound strengthened far-infrared radiation drying on potato slices, a far-infrared radiation with ultrasound assistance was applied to study the effects of different ultrasound power and far-infrared radiation drying temperature on drying process and effective moisture diffusivity values of potato. The results showed that the rise of both ultrasound power and radiation temperature could promote the mass and heat transfer in the material, shorten the drying time, and enhance the drying rate effectively of potato slices. The enhancing effect of ultrasound is weakened gradually with the decrease of water content of material and strengthened with the increase of ultrasonic power. The ultrasound strengthening phenomenon of first strengthening and then weakening with the increase of radiation temperature indicates that excessive radiation temperature will weaken the internal mass transfer effect. The effective moisture diffusion coefficient values are in the range of 1.15×10^-10～3.18×10^-10 m²/s, and the increases of both ultrasound power and radiation temperature could promote the flow and diffusion of water, and therefore, improve the corresponding D_eff values. According to the drying process of potato, three neural network models, Elman, RBF and BP, were constructed to predict the material moisture content. The results showed that all the three neural networks could achieve good prediction performance. The BP network model, which using the optimization algorithm could reach the highest fitting precision between the predicted and real values, which indicates that the model could quickly and accurately predict the moisture content of potato during ultrasonic enhanced far-infrared radiation drying process.