In recent years, there has been much research on context-aware computing, but there are not enough efforts on the correlation among contexts. The primary focus of this paper is to discover context interrelationship. In a context-aware system, there may be multiple contexts created, activated, utilized and discarded in the system. What is the influence they exert on each other? To answer this question we must discovery their inherent relationship from the time they created. In this paper we study the different kinds of interrelationship among contexts, including independence, concurrency, cascading and hierarchy. We express the different phenomena in detail and propose algorithms to deal with different problems. At last we use a case of study to illuminate our approaches.