BACKGROUND: Loosening associated with cemented polyethylene glenoid components is a major concern following total shoulder arthroplasty (TSA). The purpose of this study was to investigate the clinical and radiographic results associated with use of a novel uncemented soft-metal-backed glenoid component (SMBG), with a minimum follow-up of 2 years. MATERIALS AND METHODS: Twenty-two patients (19 women) underwent TSA using a uncemented SMBG. The mean age was 68.5 years (range, 49-84). Mean follow-up was 50 months (range, 24-89). Indications for TSA were primary osteoarthritis (10), post-traumatic osteoarthritis (8), steroid-induced avascular necrosis (2), crystalline arthropathy (1), and arthritis secondary to systemic lupus erythematodes (1). Subjective and objective parameters were assessed. Loosening and polyethylene wear were evaluated. RESULTS: Mean absolute Constant scores improved from 29.1 to 65.9 points (P < .001), age- and sex-adjusted Constant scores improved from 40.1 to 87.7% (P < .001), and subjective shoulder values improved from 35% to 75.2% (P < .001). Mean pain scores improved from 4.2 points to 13.1 (P < .001). Three cases had a fractured glenoid component. Only these 3 had a definite loosening. Polyethylene wear was found in 2 cases. CONCLUSION: Use of an uncemented SMBG component yields controversial results. Osteointegration appears possible and loosening signs have virtually not been observed. Conversely, the current implant can be associated with a high failure rate (13.6%) because of implant fractures despite short follow-up. As loosening seems absent or minimal but implant stability insufficient, design changes need to be performed and tested in view of solving the implant failure problem while preserving the actually excellent bone-implant interface characteristics.