Smart home system is an important realization of pervasive computing. Context-aware is the key technology to realize forecasting user's behavior and making the right response. Due to the context of pervasive computing with time-sensitive, the scheduling management module in the smart home system must support real-time scheduling of context-aware. In reasoning module, application queries may cause repetitive collections when context information lose efficacy. In order to solve this problem, this paper proposes a context priority scheduling algorithm based on correlation. The correlation between low-level contexts of reasoning task will be considered as the base of priority distribution. Its purpose is to speculate emergencies that about to happen next and provide the better reflect and the high efficiency in the smart home system.