Despite the great interest for the role played by resident assemblages in regulating biological invasions, few studies have assessed how these can influence the spread of exotic species that have successfully established or have included more than one trophic level. On shallow rocky reefs, we assessed how the effects of different benthic assemblages on the spread of an invasive alga, Caulerpa racemosa, are influenced by alterations in the density and species composition of the resident sea urchin assemblage. In order to simulate herbivore species loss scenarios, assemblages dominated by different morphological groups of algae (i.e. turfs or encrusting corallines) or experimentally cleared plots (i.e. bare rock) were exposed to grazing by different combinations of species (Arbacia lixula and Paracentrotus lividus) and densities (natural, −50% and −100% of natural densities) of urchins. Algal turfs and encrusting corallines generally facilitated C. racemosa. Manipulating urchins assemblages did not affect the cover and density of fronds of C. racemosa. In contrast, halving the density of P. lividus favoured the penetration of stolons of C. racemosa, consistently among algal assemblages. Other effects of urchins varied among algal assemblages, indicating interactions between trophic levels. In algal turfs, the total removal of urchins caused a decrease in the penetration of stolons, while it enhanced the length of fronds, indicating a shift in the growth form of this clonal plant and, ultimately, a depression of its spreading ability. In bare or encrusting corallines dominated surfaces, the removal of urchins had positive effects on the penetration of stolons and on the length of fronds of C. racemosa, irrespective of one or two species being manipulated and for the intensity of their removal. Our results show that tradeoffs between negative and positive effects of herbivores, varying according to relative densities of species and to the direction and strength of the effects of resident plant assemblages, can influence local rates of spread of C. racemosa. Thus, not only facilitation of exotics by natives can be key in enhancing the spread of exotic species, but it can occur between organisms at different trophic levels. Finally, our findings have important implications for the management of C. racemosa, in view of the progressive domination of shallow rocky reefs by algal turfs.