Hepatitis C virus (HCV) NS5B polymerase is a key for viral replication enzyme and a well validated drug target. In this thesis, we attempt to develop virtual screening (VS) protocols in order to identify novel inhibitors of HCV NS5B polymerase. In the first part, several screening approaches have been evaluated. To further improve VS, combining docking with filtering methods i.e. structural interaction fingerprint and random forest classification, have been investigated. In addition, we customized a simply docking strategy called ‘two sites docking’ based on an idea that a good candidate compound should bind specifically to one target. Second part, pharmacophore modeling of a potent inhibitor HCV-796, docking and molecular dynamics (MD) simulation studies were carried out. Affinity analysis of different genotypes was discussed and the suggestion for further development was proposed. Due to lack of a proofreading mechanism, viral variants make it challenging for developing effective HCV antiviral agents. We therefore employ MD simulation to examine the binding affinity of two benzimidazole inhibitors to the reported point mutants.