Bayesian structural equation model offers a methodology to incorporate researcher’s theories and prior beliefs by utilizing informative priors for parameters with strict restrictions, such as assuming no cross-loadings. This model combines the benefits of structural equation model, which accounts for measurement error, with the advantages of Bayesian estimation method, which integrates prior information. It is easier to deal with small samples and complex models. Furthermore, it provides advantages by relaxing strict assumptions found in traditional structural equation models, such as those regarding cross-loading parameters and the diagonal residual covariance matrix. Nevertheless, the utilization of the Bayesian structural equation model in the field of psychology in China remains relatively limited. The primary objective of this study is to introduce the principles, utilization of priors, assessment of model convergence, model evaluation, and other pertinent aspects of the Bayesian structural equation model. Additionally, it illustrates the application of the Bayesian structural equation model in addressing potential cross-loading parameters using the Mplus software, accompanied by an example. The ultimate aim is to enhance researchers’ understanding of the Bayesian structural equation model and its potential for resolving practical research problems.