一、社群成員間的分享行為是普遍存在。 二、就個人背景變項所造成的差異程度而言，年齡對社群成員之社群認同有顯著 的差異性。 三、就個人背景變項所造成的差異程度而言，不同使用虛擬社群時間對社群成員 之社群認同有顯著的差異性。 四、社群成員的社群認同度與其參與資訊分享行為具有顯著相關性。 五、情感、行為及認知可正向且顯著地預測社群成員分享資訊，其中以認知的預 測力最大。 六、虛擬社群facebook的成員會因為個人對其他社群成員或是社群平台的社群 認同程度不同，而導致資訊分享行為產生改變。 Nowadays, with the rapid evolution of information and network technology, users can quickly contact with each other over the network to exchange information. Virtual communities have changed the way people interact. Community members learn from each other to solve problems by sharing information and knowledge. Literature indicates that the concept of virtual community is the same as the face-to-face community. Community members often support each other to share and exchange information, and get the sense of identity and belonging to a certain degree, so that Internet users are willing to stay in these groups. This study aims to explore the relationship between community identify and knowledge sharing behavior of virtual community participants in order to make suggestions to the community managers. This is a quantitative research which targets at Facebook users who have had online information exchange experiences. A survey questionnaire was used to collect data (N=494). t-test, one-way ANOVA, Pearson correlation, and stepwise regression were employed for statistical analysis. The findings are: 1. Sharing behavior is common among community members. 2. As far as personal factors are concerned, age has significant influence on community identify of virtual community members. 3. As far as personal factors are concerned, different time using has significant influence on the community identify of virtual community members. 4. Community identify of virtual community members has significant influence on the knowledge sharing behavior. 5. Emotion, behavior and cognition can positively and significantly predict community members to share information. As a whole, cognition has the maximum of predictability. 6. Virtual community members change their information sharing behavior depending on the level of the identity to other members or community platforms.