This research aims to compare the effectiveness of online and offline learning in the Educational Statistics course for Islamic Education (PAI) students in the fourth semester of IAIN Sorong, West Papua. The research method employed is a quasi-experimental design with a pre-experimental type - the one-shot case study design, which compares two classes with two different treatments. The subjects in this research were all students of Islamic Education in the fourth semester of the 2020/2021 academic year at IAIN Sorong, namely two classes and a total of 39 students. The sample was selected using a saturated sampling technique, namely one class was to implement online teaching and learning activities and the other class was to implement offline learning. Data collection techniques used implementation observation sheets, student activity observation sheets, questionnaire sheets, and learning results test sheets. Data analysis techniques used descriptive and inferential methods. The results of the research show that: (1) the learning outcomes of students taught using online learning achieved 84.24% learning completeness with a minimum score of 70/B and an average score of 75.21; (2) student activity reached 3.12 categorized as “Good”. (3) student responses to the online learning model were positive; 4) students taught using offline learning achieved 100% learning completeness with a minimum score of 70/B and an average score of 80.79; (5) student activity reached 2.74 in the “Good” category; and (6) student responses to the offline learning model were also positive. Thus, it can be concluded that the offline learning model is more effectively applied to students in the fourth semester of PAI-B IAIN Sorong in the educational statistics course.