This paper evaluates how Vietnam's Escuela Nueva (VNEN) program, an educational reform for primary schools supported by the World Bank, affected the cognitive (mathematics and Vietnamese) and non-cognitive (socioemotional) skills of students in that country. We use propensity score matching to estimate both short-term (1-3 years) and long-term (5-7 years) average treatment effects on the treated (ATT). We find that the impacts of VNEN on students' cognitive skills are relatively small in the short-term, and that they are larger for boys, ethnic minorities, and students in Northern Vietnam. The VNEN program modestly increased primary school students' non-cognitive skills in the short-term; these impacts on non-cognitive skills are sizable and significant for ethnic minority students, although there seems to be little gender difference. The long-term impacts are less precisely estimated, but they appear to fade away, showing little or no impact of the VNEN program on cognitive skills. There is little variation of long-term impacts by gender or geographical region, although the imprecision of the estimates for ethnic minority students does not allow us to rule out large long-term impacts on cognitive skills for those students. The program's impacts on non-cognitive skills also seem to have dissipated in the long-term.