Although academic underachievement is a familiar topic in educational research, there is no consensus on how underachievers should be selected. This study aimed to compare different selection methods, absolute split method, simple difference score method, regression method and nomination, in the identification of underachievers from a sample of 126 Chinese Grade 7 students. The results indicated that the three different statistical methods showed high congruence in selecting underachievers, but nomination selected a different group of underachievers. It was suggested that, depending on research or intervention purposes, information from statistical methods and nomination might combine to provide a more comprehensive identification procedure.[Copyright of Educational Studies is the property of Routledge. Full article may be available at the publisher's website: http://dx.doi.org/10.1080/03055690120050419]