Grasses can be important regulators of species diversity and ecosystem processes in prairie systems. Although C4 grasses are usually assumed to be ecologically similar because they are in the same functional group, there may be important differences among species or between seed sources that could impact restorations. I tested whether C4 grass species identity, seed source, or grass species richness scales to influence aboveground net primary productivity (ANPP), resistance to weed invasion, or establishment of subordinate prairie species during restoration. Plots in western Iowa, United States, were planted with equal-sized transplants of one of five common grass species (Panicum virgatum L., Sorghastrum nutans (L.) Nash, Andropogon gerardii Vitman, Schizachyrium scoparium (Michx.) Nash, and Bouteloua curtipendula (Michx.) Torrey) either from local seed or from cultivar seed sources. These plots were compared to plots containing all five species in mixture and to nonplanted plots. Differences in ANPP were found among species but not between cultivars and noncultivars or between monocultures and mixtures. Panicum virgatum, S. nutans, and S. scoparium were more productive than A. gerardii and B. curtipendula. Weed invasion was much higher when plots were not planted with grasses. Schizachyrium scoparium allowed greater establishment of subordinant prairie species than all other focal grass species. There were two separate mechanisms by which grasses suppressed prairie species establishment either (1) by growing tall and capturing light or (2) by quickly filling in bare space by spreading horizontally through rhizome growth in short species. These results suggest that high ANPP can be found with noncultivar plantings during the first 2 years after planting and that subordinate species establishment is most likely when shorter bunchgrasses such as S. scoparium are dominant.