Abstract: Based on panel data from 31 provinces（Autonomous Regions and municipalities directly under the central government, excluding Hong Kong, Macao, and Taiwan）, autonomous regions, and municipalities in China from 2003 to 2022, this study analyzes the development status of Chinese-style agricultural and rural modernization． The Gini coefficient decomposition method is employed to measure regional disparities and identify their sources, kernel density estimation is used to examine dynamic evolution, and an obstacle degree model is applied to identify the primary constraining factors． Furthermore, multiple algorithms are integrated and optimized to achieve accurate prediction． The results show that,Firstly,the overall level of Chinese-style agricultural and rural modernization exhibits an upward trend, although substantial regional disparities persist; secondly,the overall disparity fluctuates over time． Intra-regional disparities increase, particularly in the eastern region, whereas inter-regional disparities decrease but remain the principal source of the overall disparity; and thirdly,during the modernization process, absolute disparities narrow in the eastern region but tend to widen in the central and western regions． A comparative evaluation of the prediction models demonstrates that the particle swarm optimization–back propagation (PSO-BP) neural network model significantly reduces prediction errors and improves goodness of fit（ R^2 ）, thereby exhibiting strong predictive performance．