Despite biological invasions are one of the main environmental problems of the twenty-first century, there is still no theoretical or empirical agreement on whether a high phylogenetic relatedness between exotic and native species positively or negatively affect invasion success. To resolve this conundrum, it has been proposed that the effect might be scale-dependent, being negative at smaller spatial scales and positive at larger scales. Here we show that this scale-dependent pattern may be a sampling artefact associated with species-area effects and a non-random pattern of species introductions. We support this conclusion with simulations and empirical data on invaded and non-invaded avian communities in regions from five continents. We further show that at smaller-scales ---where these artifacts are negligible--- invasion success generally increases with the presence of closely-related species, but that predictive accuracy largely depends on considering the influence of human-related disturbances in facilitating invasions.