The rise of antibiotic resistance has necessitated the development of alternative strategies for the treatment of infectious diseases. Antimicrobial peptides (AMPs), components of the innate immune response in various organisms, are promising next-generation drugs against bacterial infections. The ability of the medicinal leech Hirudo medicinalis to store blood for months with little change has attracted interest regarding the identification of novel AMPs in this organism. In this study, we employed computational algorithms to the medicinal leech genome assembly to identify amino acid sequences encoding potential AMPs. Then, we synthesized twelve candidate AMPs identified by the algorithms, determined their secondary structures, measured minimal inhibitory concentrations against three bacterial species ( Escherichia coli , Bacillus subtilis , and Chlamydia thrachomatis ), and assayed cytotoxic and haemolytic activities. Eight of twelve candidate AMPs possessed antimicrobial activity, and only two of them, 3967 (FRIMRILRVLKL) and 536–1 (RWRLVCFLCRRKKV), exhibited inhibition of growth of all tested bacterial species at a minimal inhibitory concentration of 10 μmol. Thus, we evidence the utility of the developed computational algorithms for the identification of AMPs with low toxicity and haemolytic activity in the medicinal leech genome assembly.