It is well-known that socio-economic background matters in determining student performance. Systematic reviews confirm that a key role in shaping this association is played by parental involvement. Not surprisingly, successful interventions in education frequently have parental engagement as a key ingredient of their protocol, and the attention paid to this factor is increasing among policymakers and evaluators. Italy is a country displaying a high level of dropping out of school, and preventing it is a central purpose for their educational system. At the same time, Italy is a country with strong family ties. This study assesses the impact of Family Group Conferences (FGC) as a preventive tool in the school setting. Researchers wondered whether FGC could be successfully used with at-risk students, to increase their parents' involvement, and to improve their well-being in school. In order to answer these questions, researchers designed and implemented a randomized controlled trial (RCT). The RCT was developed in Garbagnatese, a social district in Lombardy (Northern Italy) constituted by 7 (small to medium) cities in the province of Milan. In 2013, the intervention was offered to 17 lower secondary schools operating in the Garbagnatese district: 15 accepted the invitation. Schools were asked to anonymously refer to the central staff of the project--about 450 students (out of a population of about 4,000 individuals), 6th or 7th graders, experiencing school problems, who might benefit from a FGC. The schools provided 262 students. Before the referral and randomization processed, researchers administered a questionnaire to the student population enrolled in the 15 schools. The questionnaire contains a large set of psychological scales measuring students' self-confidence, their well-being in school, and their feeling of being supported. The same questionnaire was administered two times after the intervention, in order to have short-term and medium-term impacts. Data was collected about randomized students directly from schools, using administrative datasets. At the end of the first year, the intervention showed to be effective on four outcomes: parental involvement (as perceived by students); relations with teachers; self-efficacy about learning; and satisfaction about school experience.