Antimicrobial peptides (AMPs) are small molecules belonging to innate immunity that act against bacteria, fungi, and viruses. With the spread of bacterial strains resistant to current antibiotics, the scientific community is deeply committed to the identification and study of new molecules with putative antimicrobial activity. In this context, AMPs represent a promising alternative to overcome this problem. To date, several databases have been built up to provide information on the AMPs identified so far and their physico-chemical properties. Moreover, several tools have been developed and are available online that allow to highlight sequences with putative antimicrobial activity and predict their biological activity. These tools can also predict the secondary and tertiary structures of putative AMPs, thus allowing molecular docking studies to evaluate potential interactions with proteins/ligands. In this paper, we focused our attention on online available AMPs databases and computational tools for biological activity and tertiary structure prediction, highlighting some papers in which the computational approach was successfully used. As the identification of peptides starts from the analysis of a large amount of data, we show that bioinformatics predictions are the best starting point for the identification of new sequences of interest that can be subsequently produced and tested.