Thirty-one patients who sustained isolated 3 or 4 part dislocated proximal humerus fractures presenting to the emergency department were randomized to undergo a hemiarthroplasty (HA; n=14) or reversed shoulder arthroplasty (RSA; n=17). The primary outcome of interest was range of motion (ROM) as well as function evaluated by the Constant-Murley Shoulder (CMS) Outcome Score, were both measured at 6 weeks and 3, 6, and 12 months follow-up. Secondary outcomes of measured included pain using the visual analogue scale (VAS), Disabilities of the Arm, Shoulder and Hand (DASH), and Short Form 12 Health Survey (SF-12) (i.e. mental and physical components). VAS was measured at 6 weeks and 3, 6, and 12 months follow-up. DASH and SF-12 were evaluated at 3 and 12 months follow-up. Results revealed that CMS was statistically significantly improved at 6 and 12 months follow-up in the RSA group (p=0.01; p=0.05, respectively). ROM internal rotation was statistically significantly favored in the HA group at 6 weeks only (p<0.001) but this change was not observed afterwards at any of the other follow-ups (p>0.05). ROM anterior elevation was statistically significantly favored in the RSA group at 6 and 12 months follow-up (p=0.0020 and p=0.02, respectively) whereas ROM abduction was only found statistically significantly different at 12 months follow-up (p=0.01). Finally, VAS was statistically significantly lower in the RSA group at 12 months follow-up only (p=0.05). SF-12 and DASH scores were not statistically significantly different between the 2 groups at 3 and 12 months follow-up (p>0.05). Radiological healing of the tuberosities and heterotopic ossification were also not statistically significantly different across groups (p>0.05); 2 patients had nonunion and 1 patient had malunion in the RSA group, whilst no cases were observed in the HA group.