Objective To evaluate the cl inical results of the femoral reconstruction technique in patients with proximal femoral deformity in total hip arthroplasty. Methods Between March 2004 and June 2009, total hip arthroplasty procedures were performed on 25 patients (26 hips) with hip joint disease and proximal femoral deformity, including primary osteoarthritis of the hip joint (2 hips), developmental dysplasia of the hip (8 hips), traumatic arthritis of the hip (14 hips), and tuberculosus arthritis of the hip (2 hips). There were 10 males (10 hips) and 15 females (16 hips), with an average age of 64 years (range, 42-82 years). The disease duration was 10 months to 25 years (mean, 10.6 years). The Harris score was 44.2 ± 5.1, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score was 45.0 ± 2.7 before operation. According to Berry’s classification system of primary disease, 8 hips were classified as developmental dysplasia of hip, 7 hips as malunion of fracture, 2 hips as chronic tuberculosis, 2 hips as sl ipped femoral epiphysis, 1 hip as previous subtrochanteric osteotomy, 1 hip as previous shelf operation of the acetabulum, and 5 hips as previous internal fixation; according to the anatomic site of the deformity, there were 5 hips of greater trochanter, 10 hips of femoral neck level, 10 hips of metaphyseal level, and 1 hip of diaphysis. Results All wounds healed by first intention. Deep venous thrombosis occurred in 3 patients within 1 week, and were cured with braking and anticoagulant therapy. Bursal synovitis of great trochanter occurred in 1 patient after 6 weeks, and was eased after taking drugs for pain rel ief. All patients were followed up 1 year and 6 months to 6 years, with an average of 3 years and 3 months. The Harris score and WOMAC score at last follow-up were 88.4 ± 3.6 and 82.0 ± 5.2 respectively, showing significant differences when compared with preoperative scores (P lt; 0.05). The X-ray films at last follow-up showed good location of prosthesis and no loosening expect 1 patient who had aseptic loosening and was given revision at 8 months postopera tively. Conclusion Proximal femoral deformity classification will contribute to the choice of prosthesis and surgical strategy,thus the good results of femoral reconstruction and cl inical results would be obtained.