The problem solving ability of expert systems is greatly improved by the way of cooperation among different expert systems in a distributed expert system. However, different expert systems in a distributed expert system may use different inexact reasoning models. To achieve cooperation among these expert systems, in our methodology, the first step is to transform the uncertainty of a proposition from one inexact reasoning model to another, then the second step is to synthesize the transformed different results. In other words, the transformation is the foundation for cooperation, and so this is a very important and very interesting problem. In this paper, we give criteria to judge which heterogeneous transformations are reasonable.