The purpose of this study was to, through Classification and Regression Trees (CART) analysis, investigate and compare leaning achievements and relevant factors among high school students at different academic levels, and then establish a classification model that can predict different levels of learning achievement with different academic levels of high school students. This longitudinal study utilized the database of Taiwan Education Panel Survey, adopting waveⅠand wave Ⅲ questionnaires filled out by one selected group of students and their parents. The total sample consisted of 3022 students who were in junior high school at waveⅠand followed up when they were senior high school/vocational high school/junior college students at wave Ⅲ. The three major findings of this study are: (a) The learning achievements of high school student at different academic levels were significantly different. The learning achievements of senior high school/vocational high school/junior college students were better than that of junior high school students; (b) The factors related to learning achievement at different academic levels of high school students were different. The CART classification model for junior high school stage included three factors -- individual, family, and social network, within which 11 variables emerged. Another model for senior high school/vocational high school/junior college stage included two factors -- individual and school, in which were two variables, course type and public or private school; (c) The factors that discriminated levels of learning achievement at different academic levels of high school students varied.