In clinical research, it is important to compute an appropriate sample size. A sufficient sample size helps researchers convince the success of and intended trial. However, it costs a lot of money to recruit many subjects. So it is critical role to compute the optimal sample size n. A cross-over design(CD) is widely used for comparing effect of various treatment groups, because it has advantages of removing unexpected cause by subjects and reducing the number of subjects as subjects experience every treatment. From these advantages, we often use the CD to identify biomarkers that have differences by groups. While there are some method to determine the sample size for two groups in CDs, there is no method for comparing more than two groups. In this study, we propose a new method to compute the sample size of CDs with multiple groups as well as to be able to apply CDs with two treatment groups.