Cognitive diagnosis models（CDMs）has gained a growing interest in educational or psychological tests.Instead of considering an overall ability,a detailed latent attribute profile is considered to obtain a finer-grained evaluation and feedback.Though the popularity of CDMs,rare researches were conducted in Afghanistan.To open up the analysis and mind,CDMs are employed in this article as a very first study.Specifically,the main goal of this study is to investigate on why it gets more difficult to learn Mathematics for Afghans students,and to evaluate and determine the strengths and weaknesses of the Afghanistan students in attributes and skills.This research aims at assessing the mathematics skills of 8th grade students in Afghanistan.The response data is from the questionnaires of Trends in International Mathematics and Science Study（TIMSS）2011.Two commonly used CDMs were employed to fit the response data,including the deterministic,inputs,noisy,“and”gate（DINA）model and the deterministic,inputs,noisy,“or”gate（DINO）model.With the assistance of CDMs,we could obtain not only the item parameters,but also the skill profile for each student.Results show that the examinees do best in Number domain while do worst at Data and Chance.And in terms of model selection,DINA model fitted the data better than DINO model.