A new approach of detecting spam is expressed by this book. It is showing a new idea to detect spam mail which is the most recent threat in Internet users specially for e-mail user.By this book we proposed an efficient algorithm to detect spam. In this paper we tried to analysis every single feature of e-mail. The proposed model builds up with some parts. The algorithm designed by taking the some features of analysis step. Then after passing some real worlds spam and legitimate e-mail we tested our system and it performed a good job.The achievements of our work are-It can detect spam and ham messages separately and It works faster than other system.our proposed system model is not the absolute solution for spam detection. There are lots of possibilities to improve this model. As behavior analysis phase is the most important portion of this paper, so there has a possibility to build up a mail server that can continuously monitors users, after monitoring users it can analysis every single features at least for one year and then it can able to more successfully detects spam messages.