Ideas of limiting similarity predict two ways in which native communities could be assembled to resist invasion: match the functional traits of the native community with potential invaders or maximize the overall functional diversity of the native community. Thus, the success of restoration efforts may be improved by selecting species most likely to exclude invasive species (trait overlap) or to completely use available resources (functional diversity). To test these ideas, we varied plant functional trait diversity and nitrogen availability in a California grassland restoration experiment. Plant functional trait diversity was varied through sowing native seed mixes from the same pool of 20 native species in equal proportion (highest trait diversity) and in mixes with increased proportion of one of five life-forms: early-season annual grasses, early-season forbs, late-season perennial grasses, late-season forbs, and N-fixing annual forbs. We increased and decreased nitrogen availability with fertilizer and carbon addition, respectively. While high functional diversity did not lead to increased native cover, dominance of specific functional types which best overlapped with the invaders traits did increase the success of natives. In particular, invasion was reduced when we sowed a high proportion of early season species, possibly because this community was best able to preempt resources from the exotic species (which were generally early-active). The native community’s ability to resist invasion also increased when we reduced nitrogen availability, again suggesting a link between resource availability and invasion. Thus, we conclude that functional overlap in resource use, rather than functional diversity, may be a useful concept to guide restoration efforts.