A central goal of invasion biology is to elucidate mechanisms regulating community invasibility. Darwin’s naturalization hypothesis, one of the oldest hypotheses in invasion biology, emphasizes the importance of phylogenetic relatedness (PR) between resident and invader species for predicting invasibility. Alternatively, a recent extension of the diversity-invasibility hypothesis predicts that phylogenetic diversity (PD) of resident communities influences invasibility. Neither of these hypotheses has undergone rigorous experimental testing, and the relative contributions of PR and PD, whose effects tend to be confounded with each other, to community invasibility are unknown. Here we consider both perspectives together by independently manipulating PR and PD in laboratory bacterial assemblages. We found that while invader abundance declined significantly as PR increased, it was unaffected by PD. Likewise, we found that resident-invader functional similarity, not functional diversity of resident communities, was a significant predictor of invader abundance. Nevertheless, invader abundance was better predicted by PR than functional similarity. These results highlight the importance of considering species evolutionary relationships, especially the phylogenetic relatedness between resident and invader species, for the prediction, prevention and management of biological invasions.