The purpose of this study is to analyze the factors affecting the self-directed learning competency of high school students. To achieve this, a multilevel model analysis was conducted using data from the 2022 Gyeonggi School Education Survey. The results of the study are as follows. First, among student-level variables, educational expenses and cultural activities positively affected the dependent variable. Second, among school-level variables, the total number of students positively affected the dependent variable. Third, regarding student-level process variables, attitude toward class, peer relationships, and teacher-student relationships had a positive(+) effect on the dependent variable. When student-level process variables were introduced, the influence of educational expenses, cultural activities, and the total number of students decreased. Fourth, regarding school-level process variables, teachers' instructional competency and the degree of participation in instructional improvement activities had a positive (+) effect on the dependent variable. Fifth, the analysis of interaction effects revealed that as the total number of students increased, the influence of household educational expenses on self-directed learning competency weakened. Based on these findings, policy suggestions for improving the self-directed learning competency of high school students are presented.