Hepatitis C virus (HCV) is a major cause of liver disease worldwide. Current interferon and ribavirin (IFN/RBV) therapy is effective in 50%-60% of patients. HCV exists in infected patients as a large viral population of intra-host variants (quasispecies), which form a certain topological structure (sequence space) and may be differentially resistant to interferon treatment. We present a method for measuring differential interferon resistance of HCV quasispecies based on mathematical modeling and analysis of HCV population dynamics during the first hours of interferon therapy. The mathematical models showed that individual intra-host HCV variants have a wide range of resistance to IFN treatment in each patient. Analysis of differential IFN resistance among intra-host HCV variants allows to accurately predict outcome of IFN therapy. The models strongly suggest that resistance to interferon may vary broadly among closely related variants in infected hosts and therapy outcome may be defined by a single or a few variants irrespective of their frequency in the intra-host HCV population before treatment.