PURPOSE OF THE STUDY: We analyzed retrospectively 30 Kudo non-constrained elbow prostheses to determine: 1) functional outcome and mobility, 2) frequency of loosening and any complications. MATERIAL AND METHODS: From 1992 to 1998, 30 Kudo total elbow arthroplasties were performed in 29 patients, mean age 55 years. Mean follow-up was 36 months. These patients had severe joint disease: rheumatoid arthritis for 24, psoriatic arthritis for 2, and hemophilic arthritis for 3. The 29 patients experienced severe pain before surgery. RESULTS: At review, 21 elbows were pain free and the 9 others had only occasional pain. Among these 9 elbows, 3 exhibited a rupture of the humeral implant; one had already been revised but remained painful. One patient had a stiff painful elbow after reflex dystrophy and five others had pain but no other complication. Twenty-six patients were satisfied or very satisfied. Three patients were unsatisfied because of the humeral implant fracture. Mean mobility at last follow-up was: 128 degrees flexion, -35 degrees extension, 72 degrees pronation, and 74 degrees supination. Mean gain in flexion-extension was 15 degrees and mean gain in pronosupination was 3 degrees. Pronosupination was greater than 100 degrees except for two patients. There was one immediate post-operative dislocation with failure of prolonged orthopedic treatment after reduction; this patient underwent revision reconstruction with repair of the ulnar collateral ligaments (plasty of the medial collateral ligament with a synthetic ligament). Painful movement of the radial stump was observed with one Kudo prosthesis and required resection to achieve cure. In all, there were 3 fractures of the Kudo I prosthesis at the junction of the trochlea and the humeral stem. Among these patients, one underwent revision due to persistent pain, and two others with currently acceptable symptoms are awaiting revision. At last follow-up, we had: 1 ulnar loosening associated with cortical thinning facing the end of the ulnar implant that had migrated and showed a circular lucent line measuring > 1 mm and progressing; 9 unique ulnar lucent lines measuring<1 mm without progression at the proximal part of the implant (6 at the bone-cement interface and 3 at the bone-implant interface); 3 humeral radiolucent lines (<1 mm without progression) on the distal part of the Kudo II humeral stems corresponding to a zone without surfacing. We also observed 13 cases of incomplete ossification between the humerus and ulna and among these 13, 7 elbows had amplitudes of less than 100 degrees. DISCUSSION AND CONCLUSION: Elbow arthroplasty can restore a painless joint and maintain or improve elbow motion. The procedure is indicated when the joint disease impair daily life activities. Final mobility basically depends on the preoperative mobility. The bone stock remains the greatest problem with these resurfaced prostheses. The GUEPAR elbow prosthesis would appear to be more adapted due to the reconstruction of the trochlea. Resection of the radial head is a source of instability for elbow prostheses and should lead to the design of three-compartment prostheses.