Abstract: In the process of object-oriented change detection, the accuracy of the final result is directly related to the change threshold. Aiming at this problem, this paper presents a novel object-oriented change detection method using fuzzy comprehensive evaluation. Firstly, multi-scale segmentation is used to obtain initial objects; then, optional features for each object are chosen. Several criteria, such as objects change vector analysis, Chi-square transformation, the similarity of vector, and correlation coefficient, are treated as factors to get the “synthetic inter-layer logical values” of the fuzzy comprehensive evaluation model. The fuzzy comprehensive evaluation model is used to decide whether the target object has changed or not. Finally, the result of fuzzy comprehensive evaluation model is compared with the result of each single “inter-layer logical value” that using OTSU threshold segmentation. Based on this theory, experiments are done with SPOT5 multi-spectral remote sensing imagery. The experimental results illustrate that the model proposed can integrate the spectral and texture features and also overcome the defects caused by using single criteria. The fuzzy comprehensive evaluation model is proved to outperform other methods.