Twenty-eight patients in seventy-three patients who were diagnosed ankylosing spondylitis underwentforty-five total hip arthroplasties. Thirty-seven hips were the cementless type, and eight hips werethe cemented type. There were twenty-six males and two females. Seventeen patients were bilateralinvolvement. In forty-five total hip arthroplasties, twenty-eight hips(62%) were total ankylosis, seventeen hips(38%) were partial ankylosis. Their mean age at the operation was 39(19-63) years. Thisage seems to be a little young. The follow-up period was minimum one year and maximum seven years, theaverage being three years and four months. Six patients(21%) needed endotracheal intubation under fiberoptic laryngoscopic guide for generalanesthesia. Associated additional procedures such as Yount's operation or Campbell's operation wereperformed in fifteen hips(33%). Thirty-four(76%) hips had a excellent result; seven(15%), a goodresult; one(3%), a fair result;and three(6%), a poor result. Range of the hip motion after total hiparthroplasty was much less, but gain of the hip motion was remarkable.Pain improvement was striking in partially ankylosed hips, on the other hand motion gain was much moreimportant in completely ankylosed hips. The incidence of radiolucent line around acetabular cup andremoral stem in roentgenographic examination was 15%. One hip showed definite loosening. Preoperativespinal ankylosis took part in occurrence of radiolucent lines around prostheses. Degree of ankylosisof the hip had not specific correlation with roentgenographic radiolucent lines around prostheses. Heterotopic ossification occurred in 20%. Preoperative ankylosis of the spine and the hip, andenthesopathy had an important role to occur heterotopic Ossification. Cementless total hiparthroplasty appeared to be more superior to cemented total hip arthroplasty in ankylosing spondylitis.