Objective: Despite extensive debate, the proposed benefits and risks of video gaming in young people remain to be empirically clarified, particularly as regards an optimal level of use. Methods: In 2,442 children aged 7 to 11 years, we investigated relationships between weekly video game use, selected cognitive abilities, and conduct-related problems. A large subgroup of these children (n = 260) was further examined with magnetic resonance imaging approximately 1 year later to assess the impact of video gaming on brain structure and function. Results: Playing video games for 1 hour per week was associated with faster and more consistent psychomotor responses to visual stimulation. Remarkably, no further change in motor speed was identified in children playing >2 hours per week. By comparison, the weekly time spent gaming was steadily associated with conduct problems, peer conflicts, and reduced prosocial abilities. These negative implications were clearly visible only in children at the extreme of our game-playing distribution, with 9 hours or more of video gaming per week. At a neural level, changes associated with gaming were most evident in basal ganglia white matter and functional connectivity. Interpretation: Significantly better visuomotor skills can be seen in school children playing video games, even with relatively small amounts of use. Frequent weekly use, by contrast, was associated with conduct problems. Further studies are needed to determine whether moderate video gaming causes improved visuomotor skills and whether excessive video gaming causes conduct problems, or whether children who already have these characteristics simply play more video games.