Low student learning engagement results in suboptimal science learning outcomes. This condition indicates the need for the implementation of a learning model that is able to encourage active student involvement in the learning process. This study aims to determine the effectiveness of the Group Investigation cooperative learning model in improving science learning engagement and outcomes in the subject of force. The study used a quantitative approach with a quasi-experimental plan. The research subjects were elementary school students selected through a sampling technique based on class groups. Data collection was carried out through learning outcome tests, learning engagement questionnaires, observations of student activities, and documentation. The research instruments were test sheets, questionnaires, and observation sheets that had gone through a validation process. Data were analyzed using inferential statistical techniques to test differences in learning outcomes. The results showed that the implementation of the Group Investigation Cooperative Learning model was able to increase learning engagement and provide better learning outcomes in the cognitive, affective, and psychomotor domains compared to conventional learning. In addition, student learning engagement made a positive contribution to improving science learning outcomes. Thus, the Group Investigation learning model effectively creates more interactive and meaningful science learning. The implications of this research show that the model can be used as an alternative learning strategy to improve student activity and learning outcomes in science learning in elementary schools.