Preclinical murine infection models lack inter-laboratory uniformity, complicating result comparisons and data reproducibility. The European Innovative Medicines initiative-funded consortium (COMBINE) has developed a standardized murine neutropenic pneumonia protocol to address these concerns. While model methods have been standardized, a major obstacle to consistent results is the lack of available bacteria with defined viability and variability. Herein, we establish a diverse challenge set of Klebsiella pneumoniae and Pseudomonas aeruginosa suitable for use in the COMBINE protocol to further minimize experimental inconsistency and improve the interpretability of data generated among differing laboratories. Sixty-six K. pneumoniae and 65 P. aeruginosa were phenotypically profiled against tigecycline (K. pneumoniae only), levofloxacin, meropenem, cefiderocol and tobramycin. Fifty-nine isolates were introduced into the COMBINE model to assess the sufficiency of the starting bacterial inoculation, resultant baseline bacterial burden, achievement of ≥1 log10cfu/lung growth at 24 h, time to and percentage mortality. Forty-five isolates displaying desirable minimum inhibitory concentration profiles were subjected to replicate in vivo testing to assess target parameters. 83% of K. pneumoniae reached the prerequisite growth at 24 h using a starting bacterial burden ≥7 log10cfu/lung. P. aeruginosa isolates grew well in the model: 90% achieved the growth target with a starting bacterial burden of 6 log10cfu/lung. Mortality was negligible for K. pneumoniae but high for P. aeruginosa. Poor or inconsistent achievement of the 24 h growth target was seen in 11/59 isolates. With this diverse cache of viable isolates established in the COMBINE pneumonia model, future translational studies can be undertaken to set efficacy benchmarks among laboratories.