Abstract: A total of 1844 Betula platyphylla trees from 30 natural secondary B. platyphylla forest plots in Danqinghe Forest Farm, Yilan County, Harbin City, Heilongjiang Province were studied, to find the optimal model to predict B. platyphylla crown width from 10 commonly used basic prediction models, and reparameterize the model based on an analysis of the correlation between individual tree factors and crown width. Finally, introduce random effects at the plot level to construct a mixed-effects model for predicting the crown width of natural secondary B. platyphylla. The results showed that the Logistic model was most suitable for describing the variation in crown width of B. platyphylla, and the introduction of crown base height, height-to-diameter ratio and crown rate could improve the prediction accuracy of the model. The introduction of random effects at the plot level has improved the issue of heteroscedasticity in the data and further enhanced the predictive performance for crown width. Consequently, it is recommended to employ a nonlinear mixed-effects model for predicting the crown width of natural secondary B. platyphylla in practical production. This approach will enhance the accuracy of crown width estimation and provide a more scientifically valid theoretical foundation for the sustainable management of natural secondary B. platyphylla.