Finding strategies for decolonizing gut carriers of multidrug-resistant Escherichia coli (MDR-Ec) is a public-health priority. In this context, novel approaches should be validated in preclinical in vivo gut colonization models before being translated to humans. However, the use of mice presents limitations. Here, we used for the first time Zophobas morio larvae to design a new model of intestinal colonization (28-days duration, T28). Three hyperepidemic MDR-Ec producing extended-spectrum β-lactamases (ESBLs) or carbapenemases were administered via contaminated food to larvae for the first 7 days (T7): Ec-4901.28 (ST131, CTX-M-15), Ec-042 (ST410, OXA-181) and Ec-050 (ST167, NDM-5). Growth curve analyses showed that larvae became rapidly colonized with all strains (T7, ~106–7 CFU/mL), but bacterial load remained high after the removal of contaminated food only in Ec-4901.28 and Ec-042 (T28, ~103–4 CFU/mL). Moreover, larvae receiving a force-feeding treatment with INTESTI bacteriophage cocktail (on T7 and T10 via gauge needle) were decolonized by Ec-4901.28 (INTESTI-susceptible); however, Ec-042 and Ec-050 (INTESTI-resistant) did not. Initial microbiota (before administering contaminated food) was very rich of bacterial genera (e.g., Lactococcus, Enterococcus, Spiroplasma), but patterns were heterogeneous (Shannon diversity index: range 1.1–2.7) and diverse to each other (Bray–Curtis dissimilarity index ≥30%). However, when larvae were challenged with the MDR-Ec with or without administering bacteriophages the microbiota showed a non-significant reduction of the diversity during the 28-day experiments. In conclusion, the Z. morio larvae model promises to be a feasible and high-throughput approach to study novel gut decolonization strategies for MDR-Ec reducing the number of subsequent confirmatory mammalian experiments.