Abstract <p>Prosthetic joint infections are complex and associated with high economic and social costs. Polymicrobial cases remain poorly understood, particularly regarding bacterial interactions and antibiotic response within mixed-species biofilms. We introduce two approaches for characterizing polymicrobial biofilms: a predictive modeling equation and fatty acid profiling. Predictive modeling estimates biofilm behavior. Fatty acid profiling provided complementary information on species dominance and possible interspecies interactions. We integrate these results with fluorescence in situ hybridization and confocal microscopy to obtain a comprehensive view of biofilm structure and composition. Using this multi-modal approach, we characterize mono-, dual-, and triple-species biofilms formed by <i>Klebsiella pneumoniae</i>, <i>Pseudomonas aeruginosa</i>, and <i>Staphylococcus aureus</i>. In parallel, we evaluated common anti-biofilm antibiotics, alone and in combination, across all biofilm types. The model explained 72.9% of the variance in triple-species biofilm formation (<i>R</i>² = 0.73).<i> K. pneumoniae</i> dominated all mixed biofilms in which it was present, accounting for approximately 70–80% of the bacterial population. In <i>P. aeruginosa</i>–<i>S. aureus</i> biofilms, <i>S. aureus</i> outnumbered <i>P. aeruginosa</i> (93% vs. 7%). Antibiotic efficacy varied with biofilm composition and structure. Colistin plus levofloxacin was the most effective combination, achieving reductions of up to 7.5 log₁₀ CFU/cm². <i>S. aureus</i> drove biofilm persistence; no treatment achieved eradication in mixed biofilms containing this species, with median reductions remaining below the 5 log₁₀ CFU/cm² threshold. Confocal microscopy revealed that <i>S. aureus</i> was predominantly located in the deepest biofilm layers. These results support two complementary approaches for characterizing polymicrobial biofilms through structural, predictive, and biochemical analyses and highlight the importance of biofilm architecture and interspecies interactions when evaluating antibiotic efficacy in polymicrobial infections.</p> Key points <p>• <i>Modeling and fatty acid analysis predict polymicrobial biofilm behavior</i></p> <p>• <i>K. pneumoniae dominates biofilms, while S. aureus drives persistence</i></p> <p>• <i>Antibiotic efficacy depends on biofilm composition and structure</i></p> Graphical abstract <p></p>