Over the last two centuries, there has been a dramatic shift in the composition of grassland communities in California due to anthropogenic disturbances, the cessation of Native American burning and invasion by exotic species. As a result, the majority of grasslands in California are dominated by exotic annual grasses and forbs, with native species making up only a small fraction of the cover at most sites. These exotic dominated grasslands also tend to persist despite the cessation of the disturbances that created them (agricultural tillage, livestock grazing) rather than convert back to perennial communities suggesting that active restoration will be required to reestablish native grassland communities. My dissertation seeks to improve restoration outcomes in California grasslands by exploring the mechanisms that influence the establishment and persistence of native grassland species during restoration and how climate change is likely to interact with these mechanisms. I began with a meta-analysis of previous restoration research to understand if there were certain restoration techniques that consistently improved native plant establishment in grassland and sage scrub ecosystems across California. I categorized techniques by the constraint they addressed (dispersal, abiotic, or biotic) to determine which restoration practices are more likely to improve plant establishment. I explored the impact of these practices on whole communities and on different functional groups (grass, forb, and shrub). In total, I analyzed establishment success in 53 studies and found that in grassland and sage scrub communities addressing the dispersal filter (seed availability at the site) is the best way to improve plant establishment regardless of functional group. This suggests that most native plants targeted in restoration are seed limited and thus adding seeds or seedlings will lead to the greatest improvement in establishment compared to other restoration techniques. Native plant establishment can also be affected by the seed provenance, with many practitioners restricting seed collection to local areas near to a restoration site. This restoration technique, termed “local seed collecting”, is practiced because it is assumed that most plants are adapted to local environmental conditions. There is however, considerable debate about whether local seed collection should be the default seed sourcing strategy as the effects of climate change are increasingly considered in restoration planning. It is especially important to explore whether local seed sourcing is necessary for dominant species often used in restoration. One such species is Stipa pulchra, a commonly planted perennial grass in California projects that is extant across a wide geographic range. In my second chapter, I explore how different seed sourcing strategies affect the establishment and growth of S. pulchra. I established three common garden plots distributed across a latitudinal gradient in California, and into them I planted seeds collected from seven distinct populations of S. pulchra. I then monitored their growth and reproduction for two years. I found limited evidence that restricting seed collection to local populations of S. pulchra resulted in higher performance than using seeds from more distant populations. Instead, I found evidence to support an admixture seed sourcing approach as a way to increase resiliency to environmental variation. This would involve collecting seeds over a regional area that encompasses greater genetic diversity than a local approach. This approach would help increase the chance that restored populations persist into the future given the changes to the climate that are expected. Regardless of restoration technique however, as climate change progresses, there could be inevitable shifts in plant community composition in seasonal climate regions such as California particularly as droughts increase. In my third chapter, I describe a three-year rainfall manipulation study in which we explored how an increase in extreme drought and in the occurrence of late spring rains, two of the likely climate change scenarios for Mediterranean California, would impact the establishment and persistence of a restored native grassland community. I seeded experimental plots experiencing three climate situations with mixes of 12 native species as well as planted S. pulchra plants. After three-years, I found that we were able to restore diverse grassland communities, with both perennial and annual species, in climates that were both wetter and drier than ambient. We also found that dry conditions favored specific native species and that these species were relatively stable over time. I suggest that practitioners could leverage dry conditions to control invasive species and establish relatively stable native assemblages. Overall, my dissertation research suggests that a diversity of native grassland species can survive in a wide diversity of climates and that techniques are available to restore them. At the same time, my research also suggests that the future of grassland restoration will, in part, benefit the inclusion of climate change predictions into restoration planning. By acknowledging and addressing potential changes to the climate now, adjusting seed mixes and seed sources, restored grassland communities will be more likely to persist into the future.