The purpose of this study was to assess arthroscopic repair of rotator cuff tears at a minimum of 2 years postoperatively with both patient-derived and objective outcome measures, including the use of magnetic resonance imaging (MRI), to evaluate repair status. Evaluated were 49 shoulders in 47 consecutive patients. The American Shoulder and Elbow Surgeons score, Constant and Murley score, Simple Shoulder Test, Rowe score, Visual Analog Pain Scale, and the Medical Outcomes Study Short Form-12 Mental Component Scale all improved significantly (P < .001) between the preoperative and final follow-up evaluations. MRI found 22% of repairs had recurrent tears. The presence of a recurrent tear correlated significantly with patient age (P < .009) and extension of the tear to the infraspinatus (P < .009). Active forward flexion, abduction, external rotation, and strength in forward flexion correlated inversely with the presence of a recurrent tear (P < .05). At minimum 2-year follow-up, arthroscopic repair of rotator cuff tears produced significant improvements in both patient-derived and objectively measured variables.