Class I membrane fusion protein plays an important role in the entry of class I enveloped viruses. Peptide virus fusion inhibitors targeting the six-helix bundle structure of viral fusion protein were designed by mimicking partial sequence of the viral fusion protein. The peptides can interact with the viral fusion protein to form heterogeneous complexes, thus inhibiting the fusion between virus envelope and cell membrane. In classical approaches, the viral fusion inhibitors are designed mainly based on the primary and secondary structural information of fusion protein. However, to improve the antiviral activity, peptides should be optimized based on the tertiary structure of fusion protein. This may limit the rapid development of fusion inhibitors against emerging viruses because the tertiary structure information of these viral proteins is generally unavailable yet. In this study, we proposed to design and optimize viral fusion inhibitors based on the primary sequence, rather than the tertiary structure of the viral fusion protein using I-Mutant2.0 computer program. Using this program, several compatible and incompatible optimization sites in the sequence of a Middle East respiratory syndrome coronavirus (MERS-CoV) fusion inhibitory peptide were identified. Based on the analyzed results of MERS-CoV S protein HR2 region, several peptides with mutations in the compatible or incompatible optimization sites were designed and synthesized. It was found that modifications on compatible sites would significantly increase the inhibitory activity of the peptides, while modifications on incompatible sites resulted in decrease of the inhibitory activity of the peptides. The results confirmed the feasibility of this approach and laid the foundation for further development of novel virus fusion inhibitors.