The aim of this study was to assess the long-term results of the Kudo type-5 total elbow prosthesis and compare the results of two types of cemented ulnar components. The Kudo type-5 unlinked total elbow prosthesis (Biomet UK Ltd, Bridgend Wales) was developed in 1993. The stem of humeral component is porous-coated with a plasma spray of titanium alloy for cementless use. The ulnar component may be metal-backed with a porous-coated stem or polyethylene alone; the latter designed mainly for cement use. A metal-backed type without a porous-coated stem designed for cement use also came into being after 2003. Between 1993 and 2010, the Kudo type-5 total elbow arthroplasty was performed on 364 elbows in 274 consecutive patients with rheumatoid arthritis. The mean age of the patients at the time of the operation was 60.7 (27–86) years. Twenty elbows had Larsen grade III, 224 had grade IV, and 120 had grade V changes pre-operatively. Before the operation, 346 elbows had severe or moderate pain, 95 had gross valgus-varus instability. Clinical symptoms and revision rates were assessed 4 to 20 years (mean, 9 years) postoperatively. The mean Mayo elbow performance scores were all poor except for three initially (mean overall score, 39.5 points). The overall score had improved substantially both at the early follow-up (1 to 3 years after the operation) and the latest follow-up (4 to 20 years after the operation), with 89.5 and 84.7 points, respectively. At the latest follow-up the overall result was excellent for 185 elbows, good for 103, fair for 11 and poor for 30, with almost complete relief from pain for 298. The arc of movement had increased from a mean of 85.7 °pre-operatively to 95.1 ° post-operatively, and to 98.1 °at the latest follow-up. Spot welds around the humeral stem suggesting solid osseous integration were often seen in the elbows with cementless fixation of the porous humeral stem. At the latest follow-up, implants were removed due to infection in 3 elbows. Twenty-five elbows required ulnar component revision due to loosening of the all-polyethylene component. Two elbows required ulnar component revision due to loosening of the metal-backed component. Seven elbows required humeral component revision due to loosening of the humeral components. One elbow required revision due to dislocation. A survival analysis with revision or removal of one or both components as the end point was performed according to the Kaplan and Meier method. The overall survival rate of the prosthesis was 75.2% at 19 years. The survival of 87.0% in the metal-backed group was higher than the 74.3% in the all-polyethylene group. Loosening of the all-polyethylene ulnar component was the main reason for deterioration in the long-term outcome. We conclude that the long-term results of the Kudo type-5 total elbow arthroplasty is acceptable and cemented fixation of metal-backed ulnar component had better long-term survival than the all-polyethylene component