Results in patients undergoing total hip arthroplasty (THA) for femoral head osteonecrosis (ON) when compared with primary osteoarthritis (OA) are controversial. Different factors like age, THA type or surgical technique may affect outcome. We hypothesized that patients with ON had an increased revision rate compared with OA. We analysed clinical outcome, estimated the survival rate for revision surgery, and their possible risk factors, in two groups of patients. In this retrospective cohort analysis of our prospective database, we assessed 2464 primary THAs implanted between 1989 and 2017. Patients with OA were included in group 1, 2090 hips; and patients with ON in group 2, 374 hips. In group 2 there were more men (p<0.001), patients younger than 60 years old (p<0.001) and with greater physical activity (p<0.001). Patients with lumbar OA (p<0.001) and a radiological acetabular shape type B according to Dorr (p<0.001) were more frequent in group 1. Clinical outcome was assessed according to the Harris Hip Score and radiological analysis included postoperative acetabular and femoral component position and hip reconstruction. Kaplan-Meier survivorship analysis was used to estimate the cumulative probability of not having revision surgery for different reasons. Univariate and multivariate Cox regression models were used to assess risk factors for revision surgery. Clinical improvement was better in the ON at all intervals. There were 90 hips revised, 68 due to loosening or wear, 52 (2.5%) in group 1, and 16 (4.3%) in group 2. Overall, the survival rate for revision surgery for any cause at 22 years was 88.0 % (95% CI, 82-94) in group 1 and 84.1% (95% CI, 69 – 99) in group 2 (p=0.019). Multivariate regression analysis showed that hips with conventional polyethylene (PE), compared with highly-cross linked PEs or ceramic-on-ceramic bearings, (p=0.01, Hazard Ratio (HR): 2.12, 95% CI 1.15-3.92), and cups outside the Lewinnek´s safe zone had a higher risk for revision surgery (p<0.001, HR: 2.57, 95% CI 1.69-3.91). Modern highly-cross linked PEs and ceramic-on-ceramic bearings use, and a proper surgical technique improved revision rate in patients undergoing THA due to ON compared with OA