The grouping model of collaborative learning in the scope of physics learning is less discussed in secondary schools. This study aims to determine the effect of gendered-grouping on the quality of collaboration in Physics Science instructional and its effect on learning outcomes. This study uses an experimental approach with a posttest only group design model that involves two different characteristics of classes. The sampling of this study consisted of class VIII.1 consisting of 27 students (12 males and 15 females) as a homogeneous group, and class VIII.2 consisting of 31 students (14 male and 17 female) as a heterogeneous group taken by random cluster sampling. Data collection uses observation and tests to obtain quantitative data. Data analysis employed comparative statistics on collaboration level data and correlation statistics between collaboration level data and learning outcomes. The result shows that the quality of collaboration and student learning scores from homogeneous groups is better than heterogeneous groups. The correlation test results show that the level of collaboration of students from heterogeneous classes has a significant correlation with their learning outcomes. The correlation values the equation: y = 1,242.x + 8.892, which means it has a positive effect. It is concluded that homogeneous grouping has a positive effect on the quality of collaboration and the acquisition of student learning outcomes. Therefore, in improving the quality of learning collaboration using the SMP Science-Physics experimental technique, a strategy with gender homogeneous grouping worth application.