The increase in the number of pathogens due to fungi that are tolerant to therapies does not grow at the same speed than the advance on new antifungal drugs. In this sense, it is imperative to find anti-fungi peptides that are not detrimental to mammalian cells and have an effective toxicity to fungi. In this work, we use a method called polarity index, to identify anti-fungi peptides with an efficiency of 70 %. This method already published, initially identified selective antibacterial peptides from APD2 Database, and was characterized by developing a comprehensive analysis of the polar dynamics of a peptide from its linear sequence. Discriminating tests showed that in addition to being efficient in this identification, it was also good at rejecting other classifications of peptides found in that same database.