The invention of personal computers (PCs) brings a lot of convenience for many people¡¦s life. For a PC, its essential core is the operation system (OS) which is the most basic as well as important information system (IS). Because operation system is a requisite for each PC, many software companies have striven to develop and promote their own OSs. As well known, Microsoft is the most powerful company of OS and it dominated the market in the world. Microsoft is keeping designing new PC¡¦s OSs and promoting them with all its strength. It develops new version OS to fit the trade of information technology standard and users¡¦ requirements in specific product life cycle. In past, when Microsoft announced new OS version, they are quickly accepted and replaced to old ones. However, there is a strange situation for its latest OS Win 7 recently. Win 7 isn¡¦t quickly accepted to replace old version OS Win XP. The number of PC users whose platform is Win XP is still very large, and most of them have little willing to upgrade. It is a big problem for an OS company. However, this is also an interesting phenomenon and worth studying in academic. The issue is: why users would not like to upgrade a newer and more effective OS? This study focuses on the issue and tries to discover factors affecting upgrading intention and behavior. According to related research and actual observation, several critical constructs are applied such as switching costs, product qualities of relative advantages and over performance, and compatibilities. Moreover, environment factors like social influence and after-sales services expectations, and fashion effects. Demographic variables are included into the research model at the same time. Hypotheses are proposed after reviewing related studies and empirical survey. To verify this model and prove these hypotheses, PLS is applied to analyze and explain the result. On the other hand, discriminant analysis is also used to view the Win XP group and Win 7 group. The key discriminant function is made to distinguish and forecast these two kinds of groups. This study empirically validated and confirmed our research model by PLS and discriminant analysis. Furthermore, the relationships of factors those affecting upgrading intentions and behavior are verified and integrated. The findings are able to support OS suppliers to actually implement their product design. In academic, this study complement the field of IS switching about vertical upgrade to certain IS.