Summary Statement. Discovery system produced effective functional improvement in both primary and revision total elbow replacement. The incidence of major complications was in an acceptable range. Introduction. The search for the ideal elbow prosthesis continues as instability and loosening remain the prime reasons for total elbow replacement (TER) failure. The Discovery Elbow System (Biomet) is one of the latest generations of linked prosthesis and has been used in UK since 2003. We report outcome of TER using this system. Methods. A total of 100 TERs (75 primary, 25 revisions) were performed between 2003 and 2010. The main primary underlying pathologies for TER were advanced rheumatoid arthritis (N=58), osteoarthritis (N=35), acute fractures (N=7). There were 60 female and 40 male patients with an average age of 62 years. The outcome assessment included pain, patient satisfaction, Liverpool Elbow Score (LES), range of movement, and imaging during a mean follow-up period of 48.5 months. Major complications are also reported. Results. For the whole patient group (primary + revision), the LES was significantly (p<0.001) improved from 3.79+/−1.71 to 6.36+/−1.85There were significant improvements in elbow flexion from 100°+/−24 to 118°+17, supination from 38°+/−26 to 50°+/−25 and pronation from 48°+/−22 to 61°+/−21. Mean improvement in flexion-extension and pronation-supination arc was 20° and 25°, respectively. 64% of cases were completely pain-free and at the final follow-up (compared to 7% preoperatively). Only 6% of patients scored “Not Satisfied” at the final follow-up. LES improvement was significantly higher in the primary TER compared to revision TER (p<0.05). Imaging reviewed for 60 cases showed loosening in 4% of patients. Other main complications included deep infection (N=2), ulnar neuropathy (N=3), pre-prosthetic fracture (N=2), and prosthetic failure (N=1). Discussion. TER using the Discovery Elbow System is an effective arthroplasty in terms of functional improvement, pain relief and range of motion in both primary and revision patients. TER resulted in no/mild pain in 78% of cases. Patients undergoing Acclaim, Souter-Strathclyde, GSB III, and Coonrad-Morrey TER have been reported to have no/mild pain in 64%, 67%, 50–92% and 60–100% of cases, respectively. A 20° improvement in flexion-extension arc is comparable to that of Acclaim (23°), Souter-Strathclyde (15°), GSBIII (19–33°), and Coonrad-Morrey (17–26°) TER. An improvement of 25° in pronation-supination arc in our series is also comparable to that of 21–28° reported the Coonrad-Morrey and 27–33° for Discovery prostheses. An infection rate of 2% is lower than several other reports for GSB III TER (7–11%) and Coonrad-Morrey (6–8%). The incidence of persistent ulnar neuropathy (3%) was lower compared to GSBIII TER (11–14%), Coonrad-Morrey (12–26%), and Acclaim (8%)