Global trends show increase in microbial infections caused by pathogens resistant to the most common antibiotics. Antibiotics in development are usually tested on rapidly dividing cells in a culture medium and do not reflect complexity of infections in vivo, while testing in vivo is limited, expensive and ethically concerning. This often results in development and subsequent prescription of antibiotics only targeting infections in which pathogens are undergoing rapid cell division and in case of persistent infections like keratitis leads to poor clinical outcomes such as impaired vision or loss of an eye. In this study, we demonstrate antibiotic tolerance of Pseudomonas aeruginosa strains PAO1 and PA14 using the ex vivo porcine keratitis model in which bacterial physiology more closely mimics infections in vivo than in a culture medium. MBEC and MIC was used as a guideline to establish concentration of applied antibiotics on tissue. Infected ex vivo porcine corneas were treated with therapeutically relevant concentrations of gentamicin, ciprofloxacin, chloramphenicol, clindamycin and fusidic acid. Ciprofloxacin was the most potent across all tests demonstrating a positive correlation with MIC but not MBEC. Nonetheless, the results demonstrated that MIC and MBEC concentrations were not sufficient to clear infection even after 18 hours of continuous exposure to the tested antibiotics reflecting the need for novel antibiotics that can target the persistent subpopulation of these pathogens and the ability of the ex vivo keratitis model to be a relevant platform to identify novel antibiotics with suitable activities. There was a clear visual distinction between corneas infected with cytotoxic strain PA14 and invasive strain PAO1. In this study both strains PA14 and PAO1 showed a high level of antibiotic tolerance, which suggests that in clinical settings the treatment approach could be similar regardless of the causative strain.