Despite the increasing demand for professionals in science, technology, engineering, and mathematics (STEM), only a small portion of young people in the USA pursue a postsecondary degree in STEM. To identify the major predictors of STEM participation, this study uses a machine learning approach, a Classification and Regression Tree (CART), to analyze a wide range of individual, family, and school factors obtained from national survey data of US high school freshmen in fall 2009 who eventually enrolled in STEM college majors by 2016. The analytic results indicate that calculus credits, science identity, total STEM credits, and math achievement are the most predictive factors during the high school years of college STEM major selection. The CART-based tree also shows how these four variables interactively predict the likelihood of students enrolling in STEM college majors.