Purpose: This study aimed at investigating the effectiveness of implementing Project Based Learning (PBL) on the topic of statistics at a lower secondary school in Surabaya city, Indonesia, indicated by examining student learning outcomes, student responses, and student activity. Research Methods: A quasi experimental method was conducted over two months involving two classes of seventh grade students, consisting an experiment class (PBL) and a control class (conventional learning). Data were collected through student activity observation sheets, student responses, and pretest-posttests. Data were analyzed by employing covariance analysis (ANCOVA). Findings: Based on ANCOVA, student learning outcomes in PBL are higher than those in conventional learning. In addition, based on the results of a descriptive analysis, results of the student learning outcomes in PBL obtained more than the minimum standard score (MSC), the students' responses in learning were positive, and the students were active in the class activities. Thus, PBL is effective in statistical learning. Implications for Research and Practice: Based on the results of the research, it can be concluded that PBL was effective in statistical learning. These findings suggest that the students were enthusiastic in working the given project and actively discussed with other students in the class. We suggest teachers apply PBL on other mathematics topics so students can be enthusiastic in mathematics learning in the class. Alternatively, teachers can use PBL with technology-assisted learning to make learning more interesting for students.