Invasive plants and urban run-off constrain efforts to restore sedge meadow wetlands. We asked if native graminoids can self-restore following the removal of Typha × glauca (hybrid cattail), and if not, what limits their recovery? After we harvested Typha and depleted its rhizome starch reserves, Carex spp. expanded vegetatively (approximately 1 m over 2 years) but not by recruiting seedlings. A seedling emergence experiment showed that seed banks were depleted where Typha had eliminated the sedge meadow over a decade ago (based on aerial photo analysis). Carex seedling emergence was 75–90% lower where Carex was absent than where it remained in the plant community, and at least 17 species that were abundant 30 years ago were absent from the seed bank and extant vegetation. By varying hydroperiod, we showed that prolonged flooding prevented emergence of Carex seedlings and that a fluctuating hydroperiod reduced emergence and ultimately killed all Carex seedlings. In contrast, Typha seedlings emerged and survived regardless of hydroperiod. Thus, slow vegetative expansion by Carex, depauperate seed banks, and altered hydroperiods all constrain self-restoration. To compensate for multiple constraints on self-restoration, we recommend a long-term management approach that capitalizes on flooding and the capacity of Carex spp. to regrow vegetatively. We suggest annually harvesting swaths of Typha at the edges of clones, before or during flood events, to allow gradual, vegetative self-restoration of Carex spp.