The purpose of the this research is: (1) To know the quality of the mathematics learning outcomes that apply a authentic assessment, (2) To know the student's response to the application of an authentic assessment of mathematics learning. This study uses a quantitative method of causal-comparative type, using test instruments, questionnaires and observations. The object of this study is conditioned in 2 classes, control class (Class 8B) and experiment class (Class 8E). The results showed that the application of the authentic assessment used by teachers is good. On the basis of the data obtained, it is known that the result of learning two classes (control class and experiment class) on attitudinal aspects and skills has the same relative quality. While the learning outcomes on the knowledge aspect, the maximum value reached by the experimental class is higher than the control class (from 72.5 to 65). The effect of treatment on the experimental class is low, indicated by the value of ES (effect size) obtained equal to 0.3. The result of the study based on the student's response level shows that there are 7 students who have a low response to the learning activity (18%), 23 students have an average response to the learning activities (61%) and 8 students have a high response to learning activities. Two other categories, none of the students had very low and very high responses to learning activities (0%). This study also confirmed that Ho was rejected and Ha was accepted, so it can be concluded that  there are differences in mathematics learning outcomes among students who use authentic assessment with students who do not use authentic assessment.