Objective To explore the effectiveness and failure causes of large-head metal-on-metal total hip arthroplasty (large-head MoM THA). Methods Between March 2007 and May 2010, 159 patients (183 hips) underwent large-head MoM THA, and the clinical data were analyzed. There were 50 females (54 hips) and 109 males (129 hips) with an average age of 50 years (range, 20-78 years). Single hip was involved in 135 cases (left hip in 69 cases and right hip in 66 cases) and double hips in 24 cases. The causes included femoral head necrosis in 74 cases (93 hips), Legg-Calve-Perthes in 1 case (1 hip), osteoarthritis in 18 cases (19 hips), developmental dysplasia of the hip in 17 cases (18 hips), osteoarthritis after hip septic infection in 8 cases (8 hips), traumatic arthritis of the hip in 6 cases (6 hips), femoral neck fracture in 17 cases (17 hips), ankylosing spondylitis in 8 cases (11 hips), rheumatoid arthritis of hip in 9 cases (9 hips), and adult onset Still’s disease in 1 case (1 hip). Before operation, visual analogue scale (VAS) was 6.59±0.87; Harris score was 45.99±8.07. Results Healing of incisions by first intention was achieved, and no operative complication occurred. The patients were followed up 1.2-8.2 years (mean, 6.1 years). Implant failure was observed in 15 cases (17 hips), and the 5-year survival rate of large-head MoM THA was 91.80% (168/183). The causes of implant failure after THA were inflammatory pseudotumor in 4 cases (4 hips), acetabular aseptic loosening in 3 cases (3 hips), osteolysis in 4 cases (5 hips), acetabular aseptic loosening combined with inflammatory pseudotumor in 3 cases (3 hips), and functional disused in 1 case (2 hips). Of them, 9 cases (11 hips) did not receive revision surgery for various reasons, while 6 cases (6 hips) underwent revision surgery at 1.2-5.4 years (mean 3.7 years) after large-head MoM THA. At last follow-up, VAS and Harris score were 1.72±1.48 and 81.37±10.75 respectively, showing significant differences when compared with preoperative scores (t=–35.547，P=0.000；t=33.823，P=0.000). The function was excellent in 44 hips, good in 89 hips, fair in 33 hips, and poor in 17 hips. Conclusion Large-head MoM THA has a high revision rate during mid- and long-term follow-up because of inflammatory pseudotumor, acetabular aseptic loosening, and osteolysis. Early revision can effectively improve the function of the hip and improve patients’quality of life.