Evidence points to the possibility of promoting executive functions (EF) through school interventions. Little is known, however, about the effectiveness of this type of intervention in situations of social vulnerability. This study investigated the effectiveness of an EF intervention program applied with a sample of preschool children, in a situation of socioeconomic vulnerability, assisted by a non-governmental organization (NGO) in northeastern Brazil. Participants were 46 children (mean age = 4.67 years; SD = 0.71), divided into experimental (EG = 25) and control (CG = 21) groups. The children were assessed before and after the intervention regarding measures of EF and behavior indices. The intervention was applied over 3 months by two professionals from the NGO. There was no group x moment interaction effect on the performance measure used, however, fewer difficulties were observed posttest in the EG in relation to the CG, according to the informant-based measures. From the parents' reports, difficulties in inhibitory control (IC) decreased in the EG. From the parents' and teachers' reports, difficulties in hyperactivity, conduct problems and behavior problems (BP) total indices also decreased in the EG. A simple mediation model suggested that gains in IC mediated the intervention impact on gains in BP as observed by the parents. Interventions in EF for socioeconomically vulnerable children can minimize the deleterious effects of poverty on neurocognitive development. This study shows the applicability and effects of this type of intervention in an educational context, within the scope of a community assistance social project, bringing neuropsychology and social requirements closer together.