To realize the uncertainty analysis of the strength of SiCf/SiC ceramic matrix composite turbine outer ring, an analysis method combining neural network surrogate model and Bayesian theory was established. Firstly, modal tests of CMC turbine outer ring were carried out to obtain the modal performance. Then, an uncertainty quantification method combining neural network surrogate model and Bayesian theory was established to quantify the uncertainty of material properties based on the modal frequency test results. Finally, considering the uncertainty of elastic constants, the uncertainty analysis of the strength of CMC turbine outer ring was conducted. Results show that after considering the uncertainty of elastic constants, the maximum equivalent stress and Tsai-Wu strength factor both follow normal distributions, with coefficients of variation of 0.01 and 0.027, respectively.