The internal ribosome entry site (IRES) of hepatitis C virus (HCV) which governs the initiation of protein synthesis from viral RNA represents an ideal target for antisense approaches. Using an original bicistronic plasmid, we first established that sequence and translational activity of HCV IRESs cloned from six patients, whether responders or not to combination therapy, were conserved. We then tested the hypothesis that antisense molecules, i.e. short peptide nucleic acids (PNA), could inhibit HCV translation by binding to the highly conserved IIId or IV loop regions of the IRES. Five 6-10mer PNAs were designed. They strongly inhibit HCV IRES-driven translation in a rabbit reticulocyte lysate assay. This inhibition was highly specific since corresponding PNAs with only one mismatch were inactive. Short phosphorothioate oligonucleotides of same sequence were unable to inhibit HCV translation. PNA molecule was shown to have anti-HCV activity in Huh-7.5 cells when electroporated with a full-length HCV genome construct. Using oligonucleotide as carrier, PNA was also transfected in HCV replicon-harboring cells and in JFH1 infected Huh-7.5 cells.