Locust plagues are an ancient phenomenon, with references going back to the Old Testament. These swarming pests are notorious for their tendency to aggregate and perform long migrations, consuming vast amounts of vegetation and decimating the cultivated fields on their path. However, when population density is low, locusts will express a solitary, cryptic, non-aggregating phenotype that is not considered as an agricultural pest. Although transition of locusts from the solitary to the gregarious phase has been well studied, the shifts in the locust microbiome composition associated with this phase-transition have yet to be addressed. Here, using 16S rRNA amplicon sequencing, we compared the bacterial composition of solitary desert locusts before and after a crowding-induced phase-transition. Our findings reveal that the microbiome is altered during the phase transition. We also show that this significant change in bacterial composition includes the acquisition of a specific bacterial species - Weissella cibaria (Firmicutes), which has been previously shown to induce aggregation in cockroaches. Our findings led us to hypothesize that the locust microbiome may play a role in inducing aggregation behavior, contributing to the formation and maintenance of a swarm. Employing a mathematical model, we demonstrate the potential evolutionary advantage of inducing aggregation under various environmental conditions; and specifically, when the aggregation-inducing microbe exhibits a relatively high horizontal transmission rate. This is a first description of a previously unknown and important aspect of locust phase transition, demonstrating that the phase shift includes a shift in the gut and integument bacterial composition.