The posterosuperior rotator cuff, composed of the supraspinatus and infraspinatus tendons, is the most common site for full-thickness rotator cuff tears and represents a significant source of shoulder disability worldwide. Recognition of and classification of full-thickness tear patterns are essential in order to optimize surgical treatment and to improve prognosis. Until recently, tear patterns have been described using one- or two-dimensional classification systems. Three-dimensional pattern recognition is critical to achieving the most successful outcome possible. For more complex patterns, a combination of side-to-side stitching, margin convergence, and interval slide techniques may be needed to achieve a tension-free tendon-bone repair. Biomechanical and anatomic evidence supports the use of linked double-row repairs for most full-thickness tears. Although double-row repairs seem to result in improved structural outcomes, clinical evidence has not shown differences in outcomes scores between single-row and double-row repairs. Single-row repair may be performed in partial-thickness, small full-thickness, or very massive, immobile tears, whereas double-row repair may be performed in most other cases.