Marine microalgae are a rich source of natural products, and their amino acid-based antimicrobial agents are usually obtained by enzymatic hydrolysis, which is inefficient and limits the research on antimicrobial peptides (AMPs) from microalgae. In this study, Aureococcus anophagefferens is used as a model to predict antimicrobial peptides through high-throughput methods, and 471 putative peptides are identified based on the de novo transcriptome technique. Among them, three short peptides, P1, P6, and P7 were found to have antimicrobial activity against Escherichia coli, Staphylococcus aureus, Micro1coccus luteus, and yeast Pichia pastoris, and they showed no hemolytic activity even at higher concentrations up to 10 mg/mL. Especially P6, a 12-amino acid peptide with three positive charges, which exhibited the most significant microbicidal effect with the lowest MIC of 31.25 μg/mL against E. coli, and electron microscope observations showed the surface of P6 treated E. coli with granular protrusions and ruptures, suggesting that it likely caused cell death by directly destroying the bacterial cell membrane. This study may enrich the database of microalgal AMPs and demonstrate an efficient process for searching and validating microalgal source AMPs by combining computer analysis with bioactivity experiments.