The Galleria mellonella infection model was used to assess the in vivo efficacy of phage therapy against laboratory and clinical strains of Pseudomonas aeruginosa. In a first series of experiments, Galleria were infected with the laboratory strain P. aeruginosa PAO1 and were treated with varying multiplicity of infection (MOI) of phages either 2h post-infection (treatment) or 2h pre-infection (prevention) via injection into the haemolymph. To address the kinetics of infection, larvae were bled over a period of 24h for quantification of bacteria and phages. Survival rates at 24h when infected with 10 cells/larvae were greater in the prevention versus treatment model (47% vs. 40%, MOI=10; 47% vs. 20%, MOI=1; and 33% vs. 7%, MOI=0.1). This pattern held true when 100 cells/larvae were used (87% vs. 20%, MOI=10; 53% vs. 13%, MOI=1; 67% vs. 7%, MOI=0.1). By 24h post-infection, phages kept bacterial cell numbers in the haemolymph 1000-fold lower than in the non-treated group. In a second series of experiments using clinical strains to further validate the prevention model, phages protected Galleria when infected with both a bacteraemia (0% vs. 85%) and a cystic fibrosis (80% vs. 100%) isolate. Therefore, this study validates the use of G. mellonella as a simple, robust and cost-effective model for initial in vivo examination of P. aeruginosa-targeted phage therapy, which may be applied to other pathogens with similarly low infective doses.