This research aims to describe and explain: 1) differences in student learning outcomes; 2) differences in student learning outcomes in terms of high self-efficacy and; 3) differences in student learning outcomes in terms of low self-efficacy using a socioscientific issue-oriented project-based learning model with direct learning. This type of research is a quasi-experiment with a matching pretest-posttest control group design. The population in this study were all students in class X of SMA Negeri 2 Abiansemal for the 2022/2023 academic year, consisting of 11 classes. Research sampling was based on class-matching techniques. The data in this study was collected using a self-efficacy scale and learning outcomes tests. The data in this research are student learning outcome scores and student learning outcome scores in terms of students' high and low self-efficacy which were analyzed using the Two-Way Analysis of Covariance (ANAKOVA) test with LSD follow-up test. The results of this research show that: 1) there are differences in the learning outcomes of students who follow the socioscientific issue-oriented project-based learning model with an average learning outcome score of 84.85 in the high category and higher than the control group of 72.02 in the high category with a significance value of 0.000 <0.05; 2) there are differences in the learning outcomes of students with high self-efficacy who follow the socioscientific issue-oriented project-based learning model with an average learning outcome score of 91.76 in the very high category and higher than the control group of 81.95 in the high category with a significance 0.000 < 0.05, and; 3) there are differences in the learning outcomes of students with low self-efficacy who follow the socioscientific issue-oriented project-based learning model with an average learning outcome score of 79.41 in the high category and higher than the control group of 75.10 in the high category