AM (Automatic Merging) is a driving support system which helps drivers to merge into a traffic lane. It is required to set its performance assessment method to see whether it meets people’s driving style of each country or region. In this paper, we propose methodologies to set suitable assessment method of AM (target performances and test conditions) which can be applied in each country or region. As for target performances, suitable ones are set by studying Japanese skillful drivers’ merging behaviors on highway and on test track. As for test conditions, a new method is proposed to calculate the possibility that a merging vehicle encounters a difficult situation by analyzing traffic camera and cloud data, which allows us to set reasonable test conditions as “X%ile difficulty” of real environment. These methodologies can be applied not only in Japan but also in other countries or regions.