The present study aimed to investigate the effects of structured adapted sport participation on physical self-concept in youth with motor disabilities, examining sport-specific developmental trajectories and the mediating role of self-determined motivation. A longitudinal design with three measurement waves (T0, T1, T2) over 12 months was used. The participants were 223 individuals aged between 13 and 28, distributed across five groups: wheelchair basketball (n = 46), Paralympic swimming (n = 44), para-athletics (n = 45), adapted martial arts (n = 43), and a non-sporting control group (n = 45). Physical self-concept was assessed with the Physical Self-Description Questionnaire–Short (PSDQ-S) and motivation was measured with the Behavioural Regulation in Exercise Questionnaire–3 (BREQ-3). Mixed ANOVAs (5 × 3) showed significant main effects of sport type and time, as well as significant Sport × Time interactions across all 11 PSDQ-S subscales (all ps ≤ .01), with between-group effect sizes ranging from η²p = .13 to η²p = .29. Sport groups showed longitudinal gains of +0.3 to +0.7 points on the PSDQ-S from T0 to T2, while the control group remained stable. Each sport produced a different self-concept profile that was consistent with its specific physical demands: swimming (flexibility, aerobic endurance), wheelchair basketball (strength, coordination), para-athletics (physical activity, endurance), and martial arts (a more balanced profile). Mediation analysis showed that self-determined motivation partially mediated the relationship between sport and physical self-concept (b = 0.31, 95% CI [0.17, 0.48]), accounting for 52% of the total effect. Cluster analysis found three profiles: high integrated, specialized, and developing physical self-concept. The results of the study extend Scarpa's (2011) cross-sectional work by providing longitudinal evidence for the role of adapted sport in shaping physical self-concept, with implications for sport orientation and autonomy-supportive coaching in disability sport contexts.