The Causality Diagram, which adopts graphical expression of knowledge and direct causal intensity of causality, overcomes some shortages in Belief Network and has evolved into a mixed causality diagram methodology coped with discrete and continuous variable. But the Causality Diagram's reasoning calculation is a NP problem. It is inconvenient its popularize and application. This paper presents an approximate reasoning algorithm, and it cuts down the Causality Diagram's steps and harsh condition, it makes the complexity of the algorithm the same order of the joint events. A reasoing example justifies our algorithm.