Abstract: The fragile geological and environmental conditions in the hilly and mountainous areas of southeastern China, along with the frequent occurrence of typhoon- and heavy-rainfall-triggered geological disasters, pose significant threats to local lives and property. While landslide susceptibility assessment has been widely researched and applied, targeted studies on prediction models for typhoon-rainfall-induced landslides remain limited. To address this gap, this study takes Anxi County, Quanzhou City, Fujian Province, as the study area and analyzes 539 cases of typhoon-rainfall-triggered landslides to examine the relationship between rainfall timing, rainfall type, and landslide occurrence. Logistic regression analysis was used to identify nine rainfall-related factors and construct a probability prediction model for typhoon-rainfall-induced landslides in the region. Based on existing data, the model's performance was evaluated using ROC curves and actual landslide records. The results indicate that heavy rain, torrential rain, and extreme rainstorm events (including extraordinarily heavy rain) are the most significant triggers of typhoon-rainfall-induced landslides. The impact of typhoon rainfall on landslide stability is both synchronous and delayed. The model achieved a prediction accuracy of 88.4%, with an ROC-AUC value of 0.94,demonstrating high accuracy and effectiveness. The proposed typhoon-rainfall landslide probability prediction model provides an important reference for landslide early warning.