Predicting community susceptibility to invasion has become a priority for preserving biodiversity. We tested the hypothesis that the occurrence and abundance of the seaweed Caulerpa racemosa in the north-western (NW) Mediterranean would increase with increasing levels of human disturbance. Data from a survey encompassing areas subjected to different human influences (i.e. from urbanized to protected areas) were fitted by means of generalized linear mixed models, including descriptors of habitats and communities. The incidence of occurrence of C. racemosa was greater on urban than extra-urban or protected reefs, along the coast of Tuscany and NW Sardinia, respectively. Within the Marine Protected Area of Capraia Island (Tuscan Archipelago), the probability of detecting C. racemosa did not vary according to the degree of protection (partial versus total). Human influence was, however, a poor predictor of the seaweed cover. At the seascape level, C. racemosa was more widely spread within degraded (i.e. Posidonia oceanica dead matte or algal turfs) than in better preserved habitats (i.e. canopy-forming macroalgae or P. oceanica seagrass meadows). At a smaller spatial scale, the presence of the seaweed was positively correlated to the diversity of macroalgae and negatively to that of sessile invertebrates. These results suggest that C. racemosa can take advantage of habitat degradation. Thus, predicting invasion scenarios requires a thorough knowledge of ecosystem structure, at a hierarchy of levels of biological organization (from the landscape to the assemblage) and detailed information on the nature and intensity of sources of disturbance and spatial scales at which they operate.