Reinforced concrete (RC) beams are important components and critical bearing members in concrete structures. Owing to the complexity of shear behavior for RC beams, most current design approaches to predict shear capacity are highly empirical, which cannot physically explain the shear failure mechanism and, consequently, should only be used within the bounds of the testing regimes from which they were derived. To accommodate this issue, a mechanics-based shear model, which can explain the physical process of shear failure mechanism with a clear physical meaning, was established to quantify the shear capacity of RC beams without stirrups with any type of longitudinal reinforcement and concrete. In this study, the force on each part of the critical failure diagonal crack plane of the RC beams without web reinforcement was analyzed, and the variation of sliding strength with any given inclination angle of diagonal crack was derived. Furthermore, the inclination angle of the critical diagonal crack along which failure occurs with the shear-span ratio was determined by the correlation analysis. The model was validated using 9 test specimens with steel longitudinal reinforcement performed in this study, with a further 452 published shear tests of FRP-reinforced beams without web reinforcement. In addition, the selected design codes and available models were used to calculate the shear strength of test beams. The comparison result shows that a reduction to different degrees in both the mean and scatter of the ratio of the experimental-versus-predicted shear capacity result for the proposed mechanics-based model. Meanwhile, the proposed model can accurately reflect the influence of the main shear parameters, and the prediction accuracy shows no obvious correlation with the wide range variation of these shear parameters.