Antimicrobial peptides (AMPs) are regarded as the most promising candidates for next-generation antimicrobial agents, with amphipathicity serving as a critical factor in de novo design or systematic optimization. Previous studies rarely investigated into imperfect amphipathicity scenarios. Therefore, we developed a series of peptides exhibiting diverse imperfect amphipathicity derived from the hybrid peptide P18 to explore the effect of different imperfect amphipathicity on peptides by evaluating their antimicrobial activity and toxicity. We found that imperfectly amphipathic peptides exhibited stronger antimicrobial activity than their perfect counterparts, among them, peptides with incompletely hydrophobic surfaces and completely hydrophilic surfaces showed higher activity than those with incompletely hydrophilic surfaces and completely hydrophobic surfaces. Moreover, imperfect amphipathicity did not lead to an increase in hemolytic activity and incompletely hydrophilic surface did not lead to an increase in cytotoxicity. Notably, APH143 (LWKKFKLKKKFLWLWKKF-NH2), a completely imperfect amphipathic peptide, exhibited stronger antimicrobial activity (MIC values of 2-4 μg/mL) against common clinical pathogenic bacteria (K.pneumoniae, P.aeruginosa, A.baumannii, S.aureus) and low hemolytic activity (HC10 > 256 μg/mL), compared to other designed peptides. Subsequent experiment indicated that APH143 possessed good stability in mouse plasma, bronchoalveolar lavage fluid and different pH, temperature, salt conditions, it was also safe in mice sub-acute toxicity assay. APH143 may be a promising candidate for antimicrobial therapy. Our research deepens the understanding of designing antimicrobial peptides based on imperfect amphiphilicity and provides a novel perspective of changing degree of imperfect amphiphilicity for the design of antimicrobial peptides to overcome multi-drug resistant infection.