One of the oldest ideas in invasion biology, known as Darwin's naturalization hypothesis, suggests that introduced species are more successful in communities in which their close relatives are absent. We conducted the first experimental test of this hypothesis in laboratory bacterial communities varying in phylogenetic relatedness between resident and invading species with and without a protist bacterivore. As predicted, invasion success increased with phylogenetic distance between the invading and the resident bacterial species in both the presence and the absence of protistan bacterivory. The frequency of successful invader establishment was best explained by average phylogenetic distance between the invader and all resident species, possibly indicating limitation by the availability of the unexploited niche (i.e., organic substances in the medium capable of supporting the invader growth); invader abundance was best explained by phylogenetic distance between the invader and its nearest resident relative, possibly indicating limitation by the availability of the unexploited optimal niche (i.e., the subset of organic substances supporting the best invader growth). These results were largely driven by one resident bacterium (a subspecies of Serratia marcescens) posting the strongest resistance to the alien bacterium (another subspecies of S. marcescens). Overall, our findings support phylogenetic relatedness as a useful predictor of species invasion success.