Improving the quality of education for millions of children worldwide has become a global priority. This study presents results from the first experimental evaluation to test the impact of a universal school-based program on (1) the quality of school interactions (i.e., students' perceptions of the level of support/care and predictability/cooperation in their school and classrooms), and (2) students' subjective well-being (i.e., peer victimization and mental health problems). The study took place in the Democratic Republic of the Congo, a low-income country affected by decades of conflict. The evaluation employed a cluster-randomized trial, where the unit of randomization was clusters of two to six schools. Included in the analyses were 3,857 students in second through fourth grades, who attended sixty-three schools nested in thirty-nine clusters. After one year of partial implementation, multilevel analyses showed promising but mixed results. The program had a significant positive impact on students' perceptions of supportive and caring schools and classrooms, but a negative impact on their sense of predictability and cooperation. The program's average effect on students' subjective well-being was not statistically significant, but differential impacts were found for various subgroups of students. The paper concludes with a discussion of the implications of the study and future directions for research in this field.