Whereas a wide variety of in vitro models have been developed and validated to assess the effect of specific food ingredients on the human gut microbiome, such models have only been developed and applied to a limited extent for companion animals. Since the use of pre- and probiotics to improve gut health is an emerging research topic in the field of companion animals and as dogs are often used as laboratory animals in developing and testing of pharmaceuticals, the current study aimed to establish an adequate canine in vitro model. This consisted of a four-stage reactor composed of a stomach and small intestinal compartment followed by a proximal and distal colon. This semi-continuous gastrointestinal tract model allowed a long-term, region-dependent, and pH-controlled simulation of the colon-associated microbial community of dogs. Upon reaching a functional steady state, the simulated canine microbial community composition proved to be representative of the in vivo situation. Indeed, the predominant bacterial phyla present in the in vitro proximal and distal colon corresponded with the main bacterial phyla detected in the fecal material of the dogs, resulting in an average community composition along the simulated canine gastrointestinal tract of 50.5% Firmicutes, 34.5% Bacteroidetes, 7.4% Fusobacteria, 4.9% Actinobacteria, and 2.7% Proteobacteria. A parallel in vivo-in vitro comparison assessing the effects of fructooligosaccharides (FOS) on the canine microbial community composition showed a consistent stimulation of Lactobacillus concentrations in the in vivo fecal samples as well as in the in vitro canine gut model. Furthermore, the in vitro platform provided additional insights about the prebiotic effect of FOS supplementation of dogs, such as a reduced abundance of Megamonas spp. which are only present in very low abundance in in vivo fecal samples, indicating an interesting application potential of the developed canine in vitro model in research related to gastrointestinal health of dogs.