 This study presents a multi-label convolutional neural network CA, MLCNNCA, model that simulates the complex evolution of mixed land uses by labeling multiple land use types to each grid cell, and compares its performance with previous CA models. The MLCNNCA model was applied in Huiju, China for mixed land use simulation during 2009 2013 and 2013 to 2020, and showed satisfactory performances with an accuracy value of 0.912 for 2009 to 2013 and 0.896 for 2013 to 2020, and a hamming loss value of 0.048 for 2009 to 2013 and 0.055 for 2013 to 2020. The MLCNNCA model with the VGG-based architecture showed the best performance and outperformed previous and based CA models. Sensitivity analyses were conducted to investigate the uncertainty of the proposed model. The MLCNNCA model can provide a new tool for better simulation of fine-scale mixed land use changes and is expected to help formulate urban planning guidelines and achieve sustainable urban development. This article was authored by Shinshin Wu, Xiaoping Liu, Dachuan Zhong, and others. We are article.tv, links in the description below.