According to the recent literature, overall, results appear to be improving with elbow prosthetic surgery for severe arthritic problems, most cases being due to rheumatoid disease. However, a significant challenge remains: although data is accumulating, longer-term follow-up and greater numbers of patients need to be assessed before this area of surgical reconstruction emerges from the clinically experimental stage. It can be seen from the literature reviewed here that prosthetic loosening within bone, displacement of the articular surface of the implant, infection, and neurologic deficit, particularly of the ulnar nerve, can be unfortunate sequelae of prosthetic replacement of the elbow for severe arthritis. Osteoarthritis of the elbow is a good deal less common than rheumatoid disease of this joint, and the severe case is sometimes dealt with by prosthetic replacement; an alternative, in well selected cases, is the somewhat less challenging procedure of lower humeral fenestration arthroplasty. Revision of a failed elbow prosthesis can be a difficult challenge for the surgeon; the alternatives in this unfortunate situation are the use of an external brace, resection arthroplasty, or arthrodesis, all of them far from ideal. Many rheumatoid patients who undergo elbow prosthetic surgery have very significant shoulder involvement. Because of painful restriction of rotation of the gleno-humeral articulation, in particular, added forces are brought to bear on the elbow, making this one of the important factors in loosening of fully constrained prostheses. So far, there has been no long-term report of combined elbow and shoulder prosthetic replacement in such individuals, which is at present typically performed on separate occasions.