Galleria mellonella larvae have been used as a host model to study interactions between pathogens and hosts for several years. However, whether the model is useful to interrogate Riemerella anatipestifer infection biology remained unknown. This study aimed to exploit the potential of G. mellonella larvae and reveal their limitations as a host model for R. anatipestifer infection. G. mellonella larvae were shown to be effective for virulence evaluations of different R. anatipestifer strains. Furthermore, the virulent strain R. anatipestifer CH-1 had a stronger ability to proliferate than the attenuated strain R. anatipestifer ATCC 11845 in both G. mellonella larvae and ducklings. Unconventionally it was shown that G. mellonella larvae cannot be used to evaluate the efficacy of antimicrobials and their combinations. Additionally, it was shown that certain virulence factors, such as OmpA (B739_0861), B739_1208, B739_1343, and Wza (B739_1124), were specific only for ducklings, suggesting that G. mellonella larvae must be cautiously used to identify virulence factors of R. anatipestifer. Evaluation of heme uptake-related virulence genes, such as tonB1 and tonB2, required preincubating the strains with hemoglobin before infection of G. mellonella larvae since R. anatipestifer cannot obtain a heme source from G. mellonella larvae. In conclusion, this study revealed the applicability and limitations of G. mellonella as a model with which to study the pathogen-host interaction, particularly in the context of R. anatipestifer infection.A fter Robert Koch's work, Stanley Falkow established the molecular version of Koch's postulates, which have guided the study of the microbial virulence determinants in infectious diseases since the late 1980s (1). One of the key purposes of the molecular version of Koch's postulates is to test whether a microorganism is attenuated when a candidate virulence gene is inactivated. Thus, the use of animal models to identify the virulence factors of pathogens is indispensable. However, animal models are accompanied by ethical problems and cannot be used in large numbers for statistical analysis. Recently, nonmammalian models, including Caenorhabditis elegans, Drosophila melanogaster, Danio rerio, and larvae of the greater wax moth, Galleria mellonella, have been used for biological research (2-4). These organisms share many advantages over mammalian models, including being easy to obtain, cost-effective, and more ethically acceptable. However, the optimal temperature for C. elegans, D. melanogaster, and Danio rerio is below 28°C (5, 6). These models are defective since the optimum temperature for most mammal pathogens is 37°C. In contrast, the G. mellonella larva model is able to survive at 37°C. Moreover, the immune system of G. mellonella larvae