Multidrug resistant (MDR) Klebsiella pneumoniae (Kp) CG258 clones, refractory to routine antibiotic treatment, are considered a particularly serious clinical threat by the WHO as cause of infection outbreaks in humans worldwide [1]. Lytic bacteriophages (phages) are natural viral predators of bacteria and effective agents of bacterial clearance that may be used against MDR species (phage therapy) as alternatives or adjunctives to antibiotics [2]. However, implementation of phages as routine human therapeutics is hindered by poor translation of successful in vitro outcomes to in vivo applications, related to our yet limited understanding of complex bacteria-phage dynamics, that restricts a priori prediction of therapeutic efficacy [2,3]. In our work, we aim to explore these important potential barriers to the widespread and effective deployment of phage therapy for bacterial pathogens, using MDR Kp CG258 as our model pathogen. We have characterised a number of phages that specifically destroy the most resistant and virulent subtype of Kp CG258 (Kp ST258), and particularly the clade that dominates in Australian hospitals (ST258 clade 1) [4], and identified candidate tail fiber genes responsible for target specificity using Kp ST258 clade 1 capsular variants and porin mutants. Co-incubation in liquid culture of target Kp and phages, with different selective specificity, produced phage-resistant mutants with different morphology (mucoidy levels). Colonies were sequenced (Illumina NextSeq) and comparison with the parent strain showed diverse patterns of genomic variants (some shared, some unique), and multiple mutant types. We will next characterize attachment mechanisms for selected phage-Kp pairs in order to determine whether the primary selective pressure for the antagonistic changes (phenotype modulation/genetic modification) in response to virulent phage attack, is linked directly to phage receptor specificity, with crucial implications for the development of effective personalized phage treatments.