In this dissertation it is shown that primary school children with higher levels of physical fitness and better developed motor skills perform better at school. These relations were found to depend on the academic domain, as physical fitness predicted achievement in reading and mathematics, whereas motor skills were predictive of mathematics and spelling. Executive functions provided an explanation for the relations between physical fitness and academic achievement: fitter children had a better developed working memory, of which verbal working memory predicted performance in mathematics and spelling, and visuospatial working memory predicted mathematics specifically. Functional MRI was used to examine whether brain activation patterns could explain the relations of physical fitness and motor skills with visuospatial working memory. No support was found for this presumed explanation. Following, it was examined whether two 14-week interventions during physical education, one focused on aerobic physical activity, one on cognitively-engaging physical activity, improved children’s academic achievement. In addition, changes in brain activation patterns where examined to explain these presumed effects. Children who participated at a higher intensity level performed better in mathematics after both interventions, and in spelling after the cognitively-engaging intervention specifically. Although evidence was not overwhelming, there were some indications that physical activity brought about effects on brain activation, with different changes depending on the type of physical activity used. Providing children with more physical activity during school hours can thus have positive effects on academic achievement, as long as the intensity level and amount of cognitive engagement are taken into account.