Introduction. Advanced stage and a large area of necrotic bone are known risk factors for failure after transtrochanteric rotational osteotomy of the hip in patients with osteonecrosis. The purpose of this study was to determine whether there were other risk factors for failure of this osteotomy. Methods. One hundred and five patients (113 hips) underwent transtrochanteric anterior rotational osteotomy for femoral head osteonecrosis and were followed for a mean period of 51 months post-operatively. Radiographic failure was defined as secondary collapse or osteoarthritic change. Multivariate analysis with the use of a Cox proportional-hazards frailty model was performed to assess factors that influenced the secondary collapse and osteophyte formation. A Kaplan-Meier product-limit method was performed to estimate survival. Results. Secondary collapse occurred in twenty-seven hips (23.9%) and fourteen of these hips (12.4%) were converted to a total hip arthroplasty. At the most recent follow-up, Merle d'Aubigne and Postel hip scores ranged from 6 to 18 points (mean, 16 points). Multivariate analysis showed that the stage of necrosis (hazard ratio=3.28; 95% confidence interval=1.49-7.24), age of the patient (hazard ratio=1.08; 95% confidence interval=1.02-1.14), body mass index (hazard ratio=1.19; 95% confidence interval=1.03-1.38), and extent of necrosis (hazard ratio=1.08; 95% confidence interval=1.04-1.11) were associated with secondary collapse. Seven of eighty-six hips without collapse progressed to osteoarthritis. Survivorship with total hip arthroplasty and radiographic failure as endpoints was 63.4% (95% confidence interval=51.1%-75.7%) and survivorship with total hip arthroplasty, radiographic failure, and loss of follow-up as endpoints was 56.0% (95% confidence interval=44.6%-67.4%) at 110 months. Conclusion. Our study showed that age, body mass index, stage, and extent of osteonecrosis were determining factors for secondary collapse, subsequent unsatisfactory clinical results, and conversion to total hip arthroplasty. These factors should be considered in patient selection for the osteotomy