The enemy release hypothesis (ERH) attributes the success of some exotic plant species to reduced top-down effects of natural enemies in the non-native range relative to the native range. Many studies have tested this idea, but very few have considered the simultaneous effects of multiple kinds of enemies on more than one invasive species in both the native and non-native ranges. Here, we examined the effects of two important groups of natural enemies-insect herbivores and soil biota-on the performance ofTanacetum vulgare(native to Europe but invasive in the USA) andSolidago canadensis(native to the USA but invasive in Europe) in their native and non-native ranges, and in the presence and absence of competition. In the field, we replicated full-factorial experiments that crossed insecticide,T. vulgare-S. canadensiscompetition, and biogeographic range (Europe vs. USA) treatments. In greenhouses, we replicated full-factorial experiments that crossed soil sterilization, plant-soil feedback, and biogeographic range treatments. We evaluated the effects of experimental treatments onT. vulgareandS. canadensisbiomass. The effects of natural enemies were idiosyncratic. In the non-native range and relative to populations in the native range,T. vulgareescaped the negative effects of insect herbivores but not soil biota, depending upon the presence ofS. canadensis; andS. canadensisescaped the negative effects of soil biota but not insect herbivores, regardless of competition. Thus, biogeographic escape from natural enemies depended upon the enemies, the invader, and competition. Synthesis:By explicitly testing the ERH in terms of more than one kind of enemy, more than one invader, and more than one continent, this study enhances our nuanced perspective of how natural enemies can influence the performance of invasive species in their native and non-native ranges.