Melioidosis is a neglected tropical disease that continues unabated in many countries, particularly in Southeast Asia. There is no vaccine and antimicrobial treatment is expensive and complicated. Virulence models are important tools used to investigate genes involved in pathogenesis. Galleria mellonella is the larvae of the wax worm moth that has been used to model various infections. Based on previous studies, we attempted to establish an infection model using Burkholderia pseudomallei and Burkholderia thailandensis, a related species which is avirulent in humans. Injections of various forms of these species (fresh and frozen) were used to develop Kaplan-Meier plots. We also tested Burkholderia cepacia, Burkholderia vietnamiensis, Burkholeria ambifaria, and Burkholderia multivorans to understand how they affect the larvae. We found that larvae injected with B. pseudomallei and B. thailandensis did not accurately model the respective infections these species cause in humans, while the other non-virulent species did not produce disease, as expected. We conclude that G. mellonella is not an appropriate infection model for B. pseudomallei and B. thailandensis. Healthcare-associated infections (HAI) are on the rise, and place a heavy burden on our healthcare system each year. Disinfectants used in healthcare settings can reduce HAIs, but first must be evaluated for proper efficacy. To date there are few statistical models that are useful in comparing disinfectant test methods. We conducted a head-to-head comparison of two common test methods, suspension and surface, using Clostridium difficile spores as the test organism. A novel statistical method was developed to evaluate which test method better predicted disinfectant performance. An activated disinfectant that gradually lost activity over time was used in these evaluations. Results showed that the suspension test method was less variable, and was a better predictor of disinfectant efficacy over time.