This study aims to determine the effectiveness of the Two Stay Two Stray method for student learning outcomes, and to determine the effectiveness of the Two Stay Two Stray method for student learning outcomes in Islamic religious education (PAI) subjects. This study uses quantitative research with a quasi-experimental approach in which there are 2 classes to be studied, 1 class as a experimental class (experimental) and 1 class will be a non-experimental class (control), with the analysis of class VIII students and the sample of this study is class VIII D as an experimental class and class VIII B as a control class. The data of this study were obtained from tests, interviews, observations, documentation. How to take class data with cluster sampling technique. The data obtained can then be processed using statistical analysis. Furthermore, hypothesis testing is done by using t-test analysis of two paired t-test samples with the help of SPSS For Windows Release 16.0. The data obtained that the average learning outcomes of the control and experimental class at the time of the pretest is 71.9 and 71.62 with a difference of 0.28 and the average learning outcomes of the control and experimental class at the posttest namely 74.38 and 78.38 with the difference in number 4. So it can be said that the application of the Two Stay Two Stray method runs effectively and well proven by the value of learning outcomes between control and experimental classes. The ttable number with df of 33 is 2.03452, so tcount is greater than ttable, then it can be said that Ho is rejected and the research hypothesis is accepted. Thus it can be concluded that by applying the Two Stay Two Stray cooperative learning method student learning outcomes can be effective, the Two Stay Two Stray cooperative method in addition to being able to provide effective learning outcomes, can also make students more compact in each group, so that students in the experimental class (VIII D) in SMP Negeri 1 Salatiga can meet the KKM value.