To accurately study the risk assessment of landslide disasters, firstly, the environmental conditions of induced landslide disasters are regarded as a fuzzy system, and the landslide risk factors in the multi-level analysis system are constructed to build a multi-level fuzzy evaluation index system. Then, the cloud model theory is introduced to improve the importance scale and membership degree involved in the evaluation process, and the multi-level fuzzy comprehensive evaluation method of landslide risk improved by a cloud model is proposed. Thus, a multi-level fuzzy evaluation cloud model for evaluating landslide risk is established. Finally, using the improved cloud model method, a multistage fuzzy comprehensive evaluation of landslide risk is conducted for the K112+210~K112 +630 section of the Long Chuan to Huaiji Highway Project in Guangdong Province. The results show that the improved cloud model can solve the problem of uncertainty in the process of landslide preparation and occurrence, greatly improve the effectiveness of landslide evaluation results, and provide an effective reference for landslide disaster prevention.