This review discusses antibacterial peptides from the perspective of development into clinically useful chemotherapeutic drugs using short lactoferricin based peptides as examples. The review shows how important features for antibacterial activity can be identified and explored using the molecular properties of a range of natural and non-natural amino acids. The results have been further refined quantitatively using a "soft-modelling" approach where important structural parameters that influence the antibacterial activity of 15-residue model peptides were identified. The review describes how this knowledge is utilised to generate pharmacophores for antibacterial efficacy. These pharmacophores turn out to be surprisingly small and relatively consistent between typical Gram-negative and Gram-positive bacteria leading to the discovery of a novel class of short synthetic cationic antimicrobial peptides. These compounds are found to have high antibacterial activity against several bacterial strains that are resistant to commercial antibiotics, and are promising as future clinical candidates for treatment of infections caused by several clinically relevant pathogens.