The research aims to determine the effectiveness of Brain-based Learning on students' mathematical connection ability and self-efficacy. It is a quantitative research with an experiment method using Post-test Only Control Group Design. The samples are taken through a cluster random sampling by taking two classes, namely grade VIII E (as the experiment class given Brain-based learning treatment) and VIII F (as the control class). The research instruments are post-test and questionnaire. Based on the hypothesis test using Mann-Whitney test supported by SPSS 21.0, it is obtained that the sig value is 0.000 (sig. < alpha; (0,05)) for the post-test data, which means that the Ho is rejected. Therefore, it is found that the mathematical connection ability of the students' joining Brain-based learning is better than the mathematical connection ability of the students not joining Brain-based learning. Moreover, it is also obtained that the sig value is 0.0345 (sig. < alpha; (0,05)) for the questionnaire data, which means that the Ho is also rejected. Therefore, it is revealed that the self-efficacy of the students' joining Brain-based learning is better than the self-efficacy of the students not joining Brain-based learning. Thus, it can be concluded that Brain-based learning affects the students' mathematical connection ability and self-efficacy.