We hypothesized the seed biomass available to migrating waterfowl would be higher in spring- versus fall-flooded wetlands. To test this hypothesis we conducted an experiment using 5 pairs of wetland impoundments in northern Missouri, USA (2000–2002). We strategically assigned one impoundment of each pair to either a fall or spring treatment group. We estimated seed biomass in fall and in spring by clipping seed heads and collecting soil cores at 20 random locations within each impoundment. We placed exclosures near each fall sample site in spring-flooded impoundments to estimate seed loss from granivorous birds and rodents. Despite similar biomass in fall between treatments (P = 0.64), overwinter seed loss was greater in fall-flooded (79%; 1,324 ± 195 kg/ha) than in spring-flooded (31%; 653 ± 130 kg/ha) impoundments (P = 0.009). Considering barnyard grass or millet (Echinochloa spp.) only, seed loss was higher in fall-flooded than in spring-flooded impoundments (P = 0.05). Spring biomass estimates were similar inside versus outside exclosures (P = 0.63) indicating loss to granivorous birds and rodents was limited. Our results suggest that fall flooding reduces seed availability for spring migrating waterfowl. We recommend spring flooding be used in areas where impoundment water levels can be manipulated to increase seed availability for spring migrating waterfowl.