This study aims to see the effectiveness of the application of discovery learning model assisted by PhET simulation on student learning outcomes. The research method used is meta-analysis by calculating the effect size (ES) of each relevant article through several stages, namely formulating problems, collecting articles, grouping articles and calculating the effect size of each article. The results showed that there were 3 groupings of articles, namely based on the level of education, subjects and grade levels at the junior and senior high school levels. Effect Size based on the level of education obtained at the junior high school level was 1.517 with a very high category, and high school was 1.034 with a high category. The second grouping is based on subjects. Effect size based on subjects obtained in science subjects amounted to 1.517 with a very high category and for physics subjects amounted to 1.034 with a high category. The third grouping is based on grade level at the junior and senior high school levels. Effect size based on grade level for junior high school is obtained in grade VII with an effect size value of 2.61 with a very high category and for grade level at the high school level is obtained in grade X with an effect size value of 1.287 with a very high category. Based on meta-analysis, it is found that the discovery learning model assisted by PhET simulation can improve student learning outcomes.