The beneficial effect of achieving a sustained virological response (SVR) after antiviral treatment against hepatitis C virus is well established. However, it remains unclear whether unsuccessful treatment (non-SVR) also improves patient survival, especially in patients with advanced liver fibrosis. We retrospectively evaluated the incidence of death or liver transplantation in the 427 naive patients with a Child-Pugh score of A and advanced fibrosis newly admitted to the Hospital Beaujon between 2000 and 2010. Patients were followed for a median time of 5.5 years. The baseline characteristics of untreated (n ؍ 102) and treated (n ؍ 325) patients were largely similar, and there was no evidence of a bias of indication. Treated patients received a combination of interferon and ribavirin and had an SVR rate of 32%. The incidence of death or liver transplantation per 100 person-years was 1.00, 3.20, and 5.44 in SVR, non-SVR, and untreated patients, respectively. After adjusting for baseline characteristics, the risk of death or liver transplantation was significantly lower in SVR than in non-SVR patients and in non-SVR than in untreated patients (hazard ratios, 0.35 and 0.51, respectively; P ؍ 0.019 and 0.038, respectively). The effect of treatment in non-SVR patients was higher in patients who had a virological or a biochemical response than in those who did not have a virological or a biochemical response. The risk of death or liver transplantation was significantly lower in treated than in untreated patients. Moreover, there was a gradient of mortality between patients with SVRs, virological or biochemical responders, and untreated patients, suggesting that treatment, even in the absence of viral eradication, has a beneficial effect on survival.