The study examines how intensive hockey training is linked with spatial ability and academic performance. Participants were hockey players from top junior teams (N = 225, mean age = 14.25, all boys) and their unselected peers (N = 278, mean age = 15.47, all boys). Compared to the unselected group, hockey players showed lower results in 10 small-scale spatial tests (Cohen’s d ranging from 0.42 to 1.04), Raven’s Progressive Matrices (d = 0.41), and 12 school subjects (d for the sum of grades = 1.17). The differences in spatial ability remained significant after controlling for Raven’s (d varying from 0.26 to 1.03). The absence of spatial advantage in athletes suggests that effects of sports on cognition are complex: spatial ability facet-specific, sport-specific, professional and intensity level-specific. Moreover, these effects might be confounded by differences in academic engagement, investment of effort and psychological and physiological effects of intensive sports engagement