Traditional medicinal plants are rich reservoirs of antimicrobial agents, including antimicrobial peptides (AMPs). Herein, Amaranthus tricolor AMPs predicted in silico are identified via proteomics profiling. Bottom-up proteomics identified seven novel peptides spanning three AMP classes including lipid transfer proteins, snakins and defensins. Characterization via top-down peptidomic analysis of Atr-SN1, Atr-DEF1, and Atr-LTP1 revealed unexpected proteolytic processing and enumerated disulfide bonds. These results highlight the potential for integrating AMP prediction algorithms with complementary -omics approaches to accelerate characterization of biologically relevant AMP peptidoforms.