Personalized recommendation is an important application of context-aware system and is being used widely. However, most existed systems provide personalized information to users only based on user preferences or user activity history. Few applications consider the user?s psychology characteristics before they make recommendation. We take this characteristic into consideration, and believe the psychology knowledge is valuable to personalized recommendation. In this paper we propose a new way for personalized recommendation in context-aware system which named field dependence/independencebased recommendation. Field dependence/independence is a cognitive style of human and reflects the requirement of user about information. We compare the field dependence user with field independence user in context-aware environment, and make field dependence/independence be a characteristic parameter in personalized recommendation. We implement it and the performance seems excellent.