Ultrasound (US) and magnetic resonance imaging (MRI) of 422 cases were evaluated to compare the feasibility in diagnosing full-thickness rotator cuff tears (FTRCTs). On the basis of different US performers, they were divided into two groups: Group 1 performed by a 5-year experience technician and Group 2 performed by a 10-year experience radiologist. Sensitivity, negative predictive value (NPV), accuracy of US, and correlation between the two modalities were better in Group 2. When an expert is available, US can be used for diagnosing FTRCTs; otherwise, MRI should be performed.