The use of learning models and media that are not in accordance with the 2013 Curriculum has an impact on the low learning outcomes of students. This study examines the effectiveness of the Microsoft Sway-assisted Problem-Based Learning Model on learning outcomes in fifth-grade science at elementary school of Gugus Kartini Mejobo Kudus compared to the Direct Instruction Model. This research is included in the type of experimental research with a quasi-experimental research design in the form of a Nonequivalent Control Group Design. The population in this study were 95 fifth-grade students at elementary school of Gugus Kartini, Mejobo District, Kudus Regency using purposive sampling techniques. In this study, Elementary School 1 Payaman served as the experimental class and Elementary School 1 Gulang as the control class. Data collection techniques using test techniques (pretest and posttest); and non-test techniques with observation, interviews, documentation, and questionnaires. Data analysis techniques include initial data analysis by normality and homogeneity tests and final data analysis by t-test and N-Gain tests. The t-test results obtained a sig value of 0.000, less than 0.05. So that Ho is rejected and Ha is accepted. This is supported by the N-Gain test results for the experimental class of 0.45 and control classes of 0.35. So, it can be concluded that the Problem-Based Learning Model assisted by Microsoft Sway is more effective than the Direct Instruction Model for science learning outcomes in fifth grade elementary school.