This thesis analyzes the characteristic and regulation of anti-spam technologies. Based of these facts, this paper brings forward a collaborative anti-spam filtering model for E-mail. Our system not only defends the repeated spam mails at the router layer but also has a higher accuracy than Spam Assassin. Presents the structure of the model and give some necessary sketch maps. Explicates carefully our idea of the design and many technologies related to the model and discusses especially many key-points too. Finally, we give the experiment results.