The promotion of regular physical activity (PA) is becoming one of the main tools applied in developed countries to address health and obesity problems, particularly in view of the proven benefits of PA on a physical, psychological, and social level. Indeed, childhood and adolescence are crucial periods for an active lifestyle can be established, with the prospect of prolonging it in adulthood. The conceptual model propounded by Stodden provides a theoretical underpinning for the relationship between motor competence (MC) and PA. This study’s objective was to explore the predictive value of motor competence (MC) regarding physical activity (PA), along with the mediating role played by self-perceived motor competence (SPMC) and comparatively perceived motor competence (CPMC), with the purpose of confirming the theoretical model propounded by Stodden, as well as the relationships among variables in our own conceptual model. To this end, we tested a random sample of 925 adolescents (53.6% males; 46.3% females, age 13.75 years, SD = 1.28). Participants completed the Multidimensional Sportcomp Battery to evaluate MC and the Achievement Motivation in Physical Education test (AMPET4) to evaluate their perception of their competence; moreover, to evaluate PA, we used three indicators from the WHO Health Behaviour in Schoolchildren survey in the Spanish version. Our final theoretical model explains 19.9% of the variance of sport practice in boys and 24.2% in girls; moreover, it ascribes an important role to self-perceived motor competence (SPMC) and to comparatively perceived motor competence (CPMC) as mediating variables in the relationship between MC and PA in both sexes. Notably, we found a direct relationship between motor coordination/control tests and PA. Overall, this study underscores the degree to which adolescents’ perceived motor competence affects their actual motor competence.