Protected areas are essential for the maintenance of biodiversity, but defining criteria for prioritizing areas to conserve is not an easy task. In general, selection has been based on species richness and endemism of plants and vertebrates; however, these do not necessarily match invertebrate data, hence the need of using other groups in conservation prioritization. Moreover, species richness represents one of several biodiversity facets and does not subsume other facets such as functional and phylogenetic diversity. Restingas are coastal ecosystems within the Atlantic Forest biome, one of the World’s biodiversity hotspots. We investigated whether there is congruence between three different spider biodiversity facets: functional (FD, the variety of functional traits of species), phylogenetic (PD, the evolutionary distinctness of species), and taxonomic (TD, the number and the relative abundance of species), and whether currently protected restingas are effective in protecting these facets. We studied vegetation-living spider communities in 11 restingas along 2,000 km of the Brazilian coast. We found that no value of any biodiversity facet was higher in protected restingas compared with unprotected ones. We demonstrated low congruence between the three biodiversity facets, so that the use of TD as a surrogate of other facets is unwarranted. Whilst some protected restingas hold high values of spider TD, other still unprotected areas present high PD or FD. This result suggests that conservation efforts should be extended to every remaining restinga because they are unique sites to at least one spider biodiversity facet. In particular, we recommend three unprotected restingas as high priorities in future conservation plans based on spider diversity, which corroborate findings for plants and vertebrates in the same sites.