Abstract: Traditional computational models for determining the shear strength of shear-critical reinforced concrete (RC) column are generally deterministic models, and exhibit large fluctuation, low computational accuracy, and poor applicability, due to the fact that they do not take into account the uncertainties of geometric conformation, material properties, and external loads. In order to overcome the above limitations, a probabilistic model for shear strength of shear-critical RC column was established based on the variable angle truss-arch model and Bayesian theory. Firstly, based on the variable angle truss-arch model, an improved deterministic computational model of the shear strength of an RC column was established by taking into consideration the influence of the axial load ratio on the critical crack angle. Then, a probabilistic computational model of shear strength for shear-critical RC columns which takes into account the influence of both epistemic and aleatory uncertainties was developed by combining the Bayesian theory and the Markov Chain Monte Carlo (MCMC) method. Finally, the applicability, accuracy, and efficiency of the proposed probabilistic computational model were validated by comparing with the experimental data and existing deterministic models. The results indicate that the proposed probabilistic computational model can describe the probabilistic characteristic of shear strength of shear-critical RC column reasonably. Meanwhile, the proposed probabilistic computational model provides a benchmark to calibrate the confidence level of traditional deterministic models. Furthermore, the proposed probabilistic computational model provides an efficient way to determine the characteristic values of shear strength of shear-critical RC columns with different confidence levels.