Summary:This paper describes the design rationale of the Select Shoulder, our surgical technique, and our 6-year clinical results utilizing this trapezoid-shaped prosthesis with a proximal porous coating and a medial suture hole. Surgery was performed on 24 patients with 24 four-part fractures or fracture dislocations. The patients were not randomized. The decision to use bone cement was based on the ability (or lack thereof) to obtain a stable interference fit. In the noncemented group, the results were better and we were able to avoid the devastating complication of tuberosity nonunion.