To evaluate the effectiveness of closed wound drainage in shoulder surgery, 300 patients were enrolled in a prospective randomized study. Three operations were studied: rotator cuff repair, anterior reconstruction for instability, and arthroplasty. One hundred patients were included in each group. All patients were evaluated for wound hematoma, infection, variation in postoperative rehabilitation caused by wound problems, and length of hospital stay. No statistical difference was found between the patients whose wounds were drained and those whose wounds were not drained. This finding existed within each category. Our data do not support the routine use of closed wound drainage in elective shoulder surgery.