Civic Education (Civics) is one of the subjects that must be taught in primary school education. In the learning process, teachers still often use conventional learning models. This learning model makes students less interested and easily bored in learning activities so that student learning outcomes are less than optimal. This also happens in Civics learning on Globalization material in elementary schools. Learning on Globalization material will be more enjoyable and meaningful if students are given the freedom to choose learning topics and make use of learning resources other than teachers so that student learning outcomes will be better. Such learning can be realised by using the Group Investigation learning model. This study aims to determine the difference in learning outcomes between students who get learning using the Group Investigation learning model and the learning outcomes of students who get learning using conventional models in Civics lessons on Globalization material. The population in this study were 51 students, namely 28 students from class IVA and 23 students from class IVB. The sampling technique used was simple random sampling technique. The number of samples obtained was 48, namely 26 students from class IVA as the experimental class and 22 students from class IVB as the control class. After learning, both classes were given post-test questions. Based on the post-test scores, a hypothesis test was carried out using the Mann Whitney U Test in the SPSS 17 program because the data obtained was not normally distributed and not homogeneous. After the U test is carried out, the Asymp value is obtained. Sig. (2-tailed) 0.044. Referring to the criteria used in the research, a value of 0.044 < 0.05 means that Ho is rejected and Ha is accepted. So it can be concluded that there is a difference in the average learning outcomes between students whose learning uses the Group Investigation model and the learning outcomes of students whose learning uses the conventional model in Civics lessons on Globalization material. From the post-test data obtained, the learning outcomes of students whose learning uses the Group Investigation model are better than the learning outcomes of students whose learning uses the conventional model. Therefore, it is recommended that the Group Investigation learning model be applied and developed, especially for complex subject matter.