Abstract: The Internet has dramatically changed the relationship among people and their relationships with others. Email is the service, providing by the Internet today for its own users; this service has attracted most of the users' attention due to the low cost. Along with the numerous benefits of Email, one of the weaknesses of this service is the continuous enhanced of the received emails. The rapid expansion of this service among the Internet users has caused that some of the individuals to exploit it resulting in the spread of spam. In this paper, we introduce a new method to detect and classify the spam. We increased the precision of Email classification through FID3 decision tree and compared the results with two methods, SVM and Naïve Bayesian, by F-Measure and precision criteria; and finally succeed to make an acceptable balance between the spam detection error instead of valid email and vice versa.