Abstract: In order to break the space limitation of the traditional prediction model of urban gas pipeline failure in information collection and calculation, a prediction model of pipeline failure was built by using cloud environment. Firstly, the human-machine-environment-management model was reviewed, the factors affecting urban gas pipelines were qualitatively analyzed by fishbone diagram, and the main factors of urban gas pipeline failure were obtained using cluster analysis method. Then, the data of urban gas accident and pipeline design indexes were collected, and the key factors affecting urban gas pipeline failure were extracted by generalized grey correlation analysis method. Finally, a pipeline failure prediction model based on logistic regression algorithm was established in the cloud environment, and the feasibility of the model was verified by practical cases. The results show that the accuracy of the prediction model reached 89%, which proved that the model was suitable for failure prediction of urban gas pipelines.