The study of pathogenicity and virulence of fungal strains, in vivo in the preclinical phase, is carried out through the use of animal models belonging to various classes of mammals (rodents, leproids, etc.). Although animals are functionally more similar to humans, these studies have some limitations in terms of ethics (animal suffering), user-friendliness, cost-effectiveness, timing (physiological response time) and logistics (need for adequately equipped laboratories). A good in vivo model must possess some optimal characteristics to be used, such as rapid growth, small size and short life cycle. For this reason, insects, such as Galleria mellonella (Lepidoptera), Drosophila melanogaster (Diptera) and Bombyx mori (Lepidoptera), have been widely used as alternative non-mammalian models. Due to their simplicity of use and low cost, the larvae of G. mellonella represent an optimal model above all to evaluate the virulence of fungal pathogens and the use of antifungal treatments (either single or in combination with biologically active compounds). A further advantage is also represented by their simple neuronal system limiting the suffering of the animal itself, their ability to survive at near-body ambient temperatures as well as the expression of proteins able to recognise combined pathogens following the three R principles (replacement, refinement and reduction). This review aims to assess the validity as well as the advantages and disadvantages of replacing mammalian classes with G. mellonella as an in vivo study model for preclinical experimentation.