Abstract: Two rock burst risk index functions were built based on stress and stress change rate respectively by using a parabolic membership function model. The burst risk index with multiple factors was calculated under the maximum membership degree principle. The spatial distributions of bearing pressure and burst risk of the working face were demonstrated comprehensively and directly by cloud pictures. Finally a cloud-based risk early warning system was obtained. This system had been applied in more than 20 coal mines with risk bursts and achieved good results.