Introduction: Children and adolescents living in shelters may present with impaired motor development, cognitive function, as well as speech and understanding; psychological alterations; and hyperactivity. All of these factors may be detrimental to motor learning. Objective: To investigate motor learning in children and adolescents living in shelters, and to compare it with that of individuals living in a family context. Methods: We assessed 36 individuals who were divided into groups: an experimental group, composed of institutionalized children and adolescents (EG, n=18), and a control group (CG, n = 18) that was matched by age and sex. Motor learning was assessed using a maze test in three stages: acquisition, retention and transfer. The data were analyzed using the Shapiro Wilk, Wilcoxon, Mann Whitney, Kruskal Wallis tests and Dunn's post-test (p < 5%). Results: The EG had a longer task performance time than the CG. There was a significant reduction in task performance time between the first