While reverse shoulder arthroplasty (RSA) is not generally accepted as a successful procedure until recent years, the concept of RSA has long been proposed. Several RSA designs were originally developed for treatment of cuff tear arthropathy (CTA) in the 1970’s, which all underwent in failure. From 1987 to 1995, Paul Grammont and his team proposed successful modern reverse designs of total shoulder arthroplasty for CTA. Medializing and distalizing the center of rotation are the most distinguished principles of Grammont’s RSA. The use of RSA has been increased dramatically over the past decades. The indications for RSA have expanded to massive irrepairable rotator cuff tears, proximal humerus fractures, fracture sequelae, inflammatory arthropathy, osteoarthritis with abnormal glenoid morphology, revision cases following failed shoulder arthroplasty, and tumor reconstruction, in addition to CTA. Contraindications to RSA includes infection, complete axillary nerve palsy, neuroarthropathy, and substantial glenoid bone erosion or defects. The number of complications and reoperations of RSA has been increased significantly with the wide application of RSA. RSA complications can be classified into non-specific complications, specific complications associated with RSA and brand-related specific problems. Glenoid bone loss is not uncommon in patients undergoing RSA, which is a challenging problem. Current methods for addressing glenoid bone loss during RSA include asymmetric reaming, use of a lateralized implant, bone grafting, and use of an augmented baseplate. Three-dimensional imaging preoperative planning software and patient-specific instruments have become commercially available for guidance in the insertion of the glenoid component in RSA in recent years. Surgeons performing RSA should select proper patients, be familiar with new surgical techniques and pay attention to technical details, so that the currently high complication rate of RSA can be reduced.