Science learning has an important role for students in dealing with issues that occur in real life. One of the issues in education is the low metacognitive ability of students. So that metacognitive ability is used as a benchmark in the assessment and success of science learning. Efforts to improve metacognition skills are by applying the STEM-based jigsaw cooperative learning model. The purpose of the study was to determine the effectiveness of the application of the STEM-based jigsaw cooperative learning model on metacognitive abilities, To find out the implementation of learning, and the activities of students in using the STEM-based jigsaw cooperative learning model for science subjects for class VII SMP Ma'arif 1 Ponorogo. The research method is quantitative research. This type of research is quasi-experimental with a nonequivalent control group design. The research instrument was an observation sheet on the implementation of learning, student activities and metacognitive ability test questions. The data analysis technique used statistical inference in the form of t-test and N-Gain test. The results of the t-test research can be concluded that H0 is rejected and H1 is accepted so that there are differences in metacognitive abilities in the experimental and control classes, then the N-Gain results show that the average of the experimental class is higher than the control class, which is 59.0817. Then the results of the implementation of learning and student activities run smoothly with a very good category. So that the STEM-based jigsaw cooperative learning model is influential and quite effective in improving metacognition abilities.