The implementation of this research is based on the problems found at the time of observation, namely the difficulty of students learning, especially in science learning that has an impact on student learning outcomes and learning that is still teacher-centered that causes students to be passive. Based on these problems, this study aims to improve the learning outcomes of science students in grade V elementary school. This type of research is a quasi-experimental design with Nonequivalent Control Group Design. The research design used is the post-test only control group design. The population in this study amounted to 147 students consisting of four classes. The determination of the sample is done by random sampling in the form of a draw. The number of samples was 72 students, namely the experimental class totaling 36 students and the control class totaling 36 students. The students' natural science post-test data collection is done by using multiple-choice objective test methods. Then, the t-test formula was analyzed. From the results of the analysis, it was shown that t-count = 3,920 was obtained and compared with t-table at the 5% significance level and dk = 36 + 36-2 = 70 so that the price of t-table = 1.99444 was obtained. Based on these results, there are significant differences in science learning outcomes between the classes given the treatment of the TSTS learning model aided by the mind mapping method and the classes given the treatment of conventional learning models. The results of the calculations obtained the experimental class has a higher mean than the control class, which is 79.06>68.75. Based on this, the TSTS learning model is assisted by an effective mind mapping method to the science learning outcomes of fifth-grade students of elementary school.