We conducted a randomized controlled trial to evaluate the impact of a summer learning program for vulnerable students across ten cities in Italy (N=1,038). The program had two components: educational workshops in small groups (88 hours) and personalized tutoring (12 hours). Results indicate significant improvements in reading comprehension and marginally in grammar. Improvements in arithmetic and geometry are smaller albeit significant when aggregated into a single mathematics score. Effects were most pronounced among primary school students and among students with special needs or from vulnerable environments. The program compensated for summer learning loss, as treatment group students returned to school in September with higher learning levels than before the summer, while the control group experienced learning setbacks, predominantly in mathematics. While the study clearly shows that students start the new year with a higher level of competencies, it does not definitively establish the lasting impact of these effects. An explorative analysis of noncognitive skills provides conflicting insights: an increase in students' interest in acquiring new competencies suggests potential enduring effects, but the emergence of dissatisfaction with traditional school activities and heightened school-related stress raises concerns about reduced engagement with conventional schooling.