Restoring roadsides with native plants can have significant benefits in erosion control, aesthetic interest, and habitat for declining pollinators. However, native plant seed mixes can be over three times the cost of more conventional plantings, and it is unclear how well native plants establish in the compacted and chemically unique soils of roadsides. We collected data to inform roadside manager decisions related to the establishment of native plants in roadsides, and to what extent they support pollinators. In this study, we revisited roadside sites that had been seeded with native versus non-native seed mixes, to assess the plant and pollinator communities 2-20 years after planting. This presentation reviews our primary findings and discusses implications for roadside revegetation and management. We found that sites seeded with native plants have significantly higher native flowering forbs than sites seeded with non-natives. Additionally, native plants do not seem to be colonizing roadsides from elsewhere: if managers want native plants in roadsides, they need to be seeding them in. Native flowering plants decrease in abundance over time since planting, as grasses become more dominant. Pollinator abundance in roadsides was primarily related to flowering plant abundance, regardless of whether plants were native. However, the diversity of butterfly and bee communities was positively related to the diversity of the flowering plant community. A parallel study on the landscape context of roadside habitat further showed that roadsides with more adjacent land designated as “bee friendly” had a greater abundance and diversity of bumblebees. Finally, we merge our data on plant establishment and pollinator preferences to consider which native plant species may represent the largest payoff for roadside habitat for pollinators. We expect these results to inform roadside revegetation efforts across the upper Midwest, helping to optimize the investment in roadside native vegetation.