Thirty-seven patients with 20 cemented Thompson-Richards prostheses and 19 cementless S.T.A.R. prostheses (2 bilateral cases) were followed up after 1-12 years. Rheumatoid arthritis was the main diagnosis in both populations, with females dominating. The investigation was based on the Kofoed ankle score. At follow-up the total scoring improved to 86.9 pts. in S.T.A.R. and to 77.7 pts. in T.R.P. replacement. The radiological examination showed a high rate of radiolucency for the tibial component (53.3%) in cemented T.R.P.; subsidence of talar component was seen in 3 cases with T.R.P. In cementless S.T.A.R. prothesis only 3 cases showed small radiolucent lines of the flat tibial component. Talar subsidence was not seen at all. In T.R.P. we had two revisions due to prothesis loosening and one maleollar fracture, giving a cumulative estimated survival rate of 87% at 12 years. In the S.T.A.R. prosthesis group two revisions had to be performed because of one meniscal breakage and correction of meniscal height. The estimated survival rate at 6 years was 94.3%.