With the development of drilling technology and reinforced exploration and exploitation of unconventional reservoirs, there has been a great increase of complex wells. Meanwhile, however, consequent casing wear is and will continue to be a serious problem that causes enormous economic losses and many safety issues. The purpose of this paper is to find out the mechanism of casing wear and establish casing wear prediction model. Casing wear experiment was carried out to study the effect of contact force, rotation speed, and casing grade on wear depth. Meanwhile, wear coefficients under different working conditions were obtained through the normalizing of data. With the extensive research of downhole drag and torque calculation method, a contact force calculation model was established. Through the combination of crescent-shaped model and wear-efficiency model, the past complicated casing wear prediction models and confusing empirical formulae were greatly simplified. Therefore, the wear volume and depth of the casing string can be accurately predicted. Finally, a prediction software was developed to predict downhole casing wear of oil and gas wells. Comparison with the field data confirmed that the established model and software had enough accuracy to help predict and analyze casing wear at field.