A total of 124 elderly patients with displaced displaced proximal humeral fractures (PHFs) were randomized to receive a reverse shoulder arthroplasty (RA) (n=64) or plate fixation (n=60). The primary outcome of interest was the Constant score at 2-years follow-up. The secondary outcomes of interest were the Oxford Shoulder Score and radiological outcomes. The results demonstrated that a statistically significant difference in Constant scores at 2 years, favour of the reverse shoulder arthroplasty group. Stratified by OTA/AO fracture classification, Constant scores were statistically significantly in favour of the reverse arthroplasty group in type C2 proximal humeral fractures. However in type B2 proximal humeral fractures, no statistically significant difference in Constant scores was observed between the reverse arthroplasty and plate fixation groups.