OBJECTIVE: To investigate whether adolescent elite female and male handball players with shoulder muscle weakness, deficits in shoulder rotation range of motion (ROM) or in joint position sense (JPS), or scapular dyskinesis in the preseason had a higher rate of new shoulder injuries compared to players without these characteristics. DESIGN: Prospective cohort study. METHODS: We studied 344 uninjured players (452 player-seasons, 50% female). We measured their shoulder strength in isometric external rotation (IER), isometric internal rotation (IIR), isometric abduction, and eccentric external rotation, as well as their shoulder ROM, JPS, and scapular dyskinesis, during the preseason. Players were monitored weekly regarding match and training hours and shoulder injuries during 1 or 2 seasons. We used multivariable Cox proportional hazard models to calculate hazard rate ratios related to the first injury and 95% confidence intervals (CIs). RESULTS: During 2 seasons, the participants reported 48 new shoulder injuries. In female players, the hazard ratio was 2.37 (95% CI: 103, 5.44) for IER weakness and 2.44 (95% CI: 106, 5.61) for IIR weakness. The hazard rate ratio was 0.85 (95% CI: 0.39,183) for an IER/IIR ratio of less than 0.75 and 1.53 (95% CI: 0.36, 6.52) for scapular dyskinesis. In male players, the hazard rate ratio was 1.02 (95% CI: 0.44, 2.36) for ER weakness, 0.74 (95% CI: 0.31,1.75) for IIR weakness, 2.0 (95% CI: 0.68, 5.92) for an IER/IIR rati of less than 0.75, and 3.43 (95% CI: 1.49, 7.92) for scapular dyskinesis. There were no associations between new shoulder injuries and deficits in ROM or JPS. CONCLUSION: In adolescent elite handball, male players with preseason scapular dyskinesis and female players with preseason IIR or IER shoulder weakness had an increased shoulder injury rate.