Introduction: Total scapular resection causes a significant functional loss because of the sacrifice of the glenoid, which serves as a stable base for shoulder motion. The authors analyze their experience with two types of reconstructions following total scapular resection; suspension of the humeral head from the clavicle without endoprosthetic reconstruction of the scapula and endoprosthetic scapular reconstruction. Materials and Methods: Between 1979 and 1997, the authors treated 23 patients with scapular tumors that required total scapular resection. Patients were diagnosed with 14 bone and 9 soft-tissue tumors. Resection included total scapulectomy in 12 patients and enbloc resection of the scapula and humeral head in 11 patients. Reconstruction: All eleven patients who had resection of their humeral head underwent reconstruction of the humerus with endoprosthesis. Scapular endoprosthesis was further installed in 7 patients and suspension of the humeral head from the clavicle with a Dacron tape was performed in 16 patients (Suspension of the prosthetic humeral head from the clavicle – 4 patients; suspension of the native humeral head from the clavicle – 12 patients). Endoprosthetic reconstruction of the scapula was feasible only when the periscapular musculature was sufficient for endoprosthetic attachment and coverage. The scapular prosthesis was attached to the prosthetic humeral head with a Goretex® sleeve, which served as an artificial joint capsule. All patients were followed for a minimum of 2 years; follow-up included physical examination, radiological evaluation and functional evaluation according to the American Musculoskeletal Tumor Society system. Results: Elbow range-of-motion and hand dexterity were similar in the two groups of patients. However, compared with patients who undergone humeral suspension, those who had scapular endoprosthesis had better abduction (60°–90° vs. 10°–20°) of the shoulder joint. Moreover, these patients had better cosmetic appearance of the shoulder girdle. There were no deep wound infections, prosthetic failures, or secondary amputations. Overall, 6 patients who had scapular prosthesis (86%) and 10 patients who had humeral suspension (62%) had a good-to-excellent functional outcome. Conclusions: The number of patients who underwent a scapular endoprosthetic reconstruction is small and does not allow a valid statistical analysis; however, the authors feel that scapular endoprosthesis reconstruction is associated with better functional and cosmetic outcomes, when compared to humeral suspension. The authors recommend reconstruction of the scapula with endoprosthesis when periscapular musculature, remaining after tumor resection allows attachment and coverage of the prosthesis