This research systematically examines the experimental articles on STEM education (science, technology, engineering, and mathematics education). In this direction, experimental articles with the Social Sciences Citation Index on the Web of Science were analyzed within the scope of research questions, taking into account the guidelines of PRISMA. The research was carried out using the systematic review method. In the first search made in this line, the results listed 17,482 studies. Then, this number was reduced to 34 by applying the inclusion and exclusion criteria. Finally, the studies were examined in detail, and 12 articles that were not suitable were excluded from the scope. Thus, this research was carried out on 22 articles. In line with research problems, articles were categorized by publication year, country, method, study group and number, data collection tools, data analysis methods, topics covered, and experimental methods. The research results infer that the number of articles increases every year, and Turkey and China are the pioneering countries in this sense. It is noteworthy that mixed methods and quasi-experimental methods are generally preferred in the literature. Contrary to the literature, another result obtained from the research is that teachers are preferred in variable numbers as the study group. Scales, tests, and interviews are the most frequently used measurement tools by researchers. Besides, t-tests and content analysis are used in data analysis. Additionally, various topics are covered in the articles. One of the results obtained from the research is that the effect of STEM education on various skills and achievements is examined more. In line with these results, some suggestions were made by discussing the studies in the literature.