The study aimed to investigate the application of the Adolescent Behavior Short-form Scale (ABSS) to serve the purpose of screen in the multi-step procedure to identic the students with emotional and behavior disorder (EBD). To meet the needs of screen, the ABSS were condensed from the teacher-and peer- rating Adolescent Social Behavior Scale (ASBS) by the BIG one parameter analysis. The items of the original teacher-and peer-rating ASBS were ll9 and 113 individually; however, the teacher-and peer rating ABSS were formed as 21 and 14 items with similar content reliabilities with the original ASBS (i.e. the alpha coefficient were .935 and .896). Similar curves in the test information curve and the test standard error between the ABSS and the ASBS were founded, which meant the ABSS provided similar quality of assessment as the ASBS. The percentile of 25 (score of 82 in teacher rating scale and 62 in peer rating scale) to serve as the cut-off scores for screening in the ABSS were also examined. The ability value of this cut-off score in the teacher rating scale is l.25 and the reliability is .88. And .57 and .90 in the cut-off score of the peer rating scale. According to the result of the ANOVA and the identification percentage of the target groups, either of short-forms ABSS can screen high-risk EBD students satisfactorily.