Thirty patients undergoing total shoulder arthroplasty were randomized to receive a continuous interscalene brachial plexus block (n=15) or a high-thoracic erector spinae plane block (n=15) for the control of post-operative shoulder pain. The primary outcome of interest was the incidence of hemidiaphragmatic paralysis. Additional outcomes of interest included incentive spirometry volume, sensory and motor assessments, pain and opioid consumption, the incidence of adverse events, and patient satisfaction. Patients who received a high-thoracic erector spinae plane block had a significantly lower incidence of hemidiaphragm paresis compared to the interscalene block group. Patients who received a high-thoracic erector spinae block also reported significantly smaller loss of incentive spirometry volume, motor function, and sensory function. Cumulative opioid consumption on post-operative day 0 was, however, significantly higher in the high-thoracic erector spinae block group.