In coastal Louisiana many restoration projects are approved based on assumed regeneration of submerged aquatic species (SAV) in shallow marsh interior ponds. In this study, we estimated seed bank size and composition of shallow water areas in oligohaline fringing and restored (terraced) marsh, and a freshwater managed (impounded) marsh, located in Sabine NWR, LA, using the sieving method. For the same marshes, we also provided an estimate of the readily germinable fraction of the seed bank using the germination method. Sieving results indicated that restored marsh edges had very low seed densities (5034 seeds/m²) compared to fringing marsh (331,185 seeds/m²), although species composition was similar. Managed freshwater marsh ponds had more diverse seed banks and mid-range seed densities (80,500 seeds/m²). Viability estimates of dominant species in the seed bank reduced seed density estimates at all sites by as much as 10 fold (fringing marsh: 36,185 seeds/m²; restored marsh: 859 seeds/m²; managed marsh 44,388 seeds/m²) suggesting that a correction factor should be applied to future seed density estimates in this region. Seedling emergence was significantly higher in the managed marsh under drawdown conditions (> 2500 seedlings/m²) as compared to flooded conditions (< 500 seedlings/m²; ANOVA, p = 0.0001). Seedling emergence in oligohaline marsh was significantly affected by salinity and management (fringing, restored) (ANOVA, p = 0.0186). Fringing marsh at 0 g/L had the highest seedling emergence (> 500 seedlings/m²). At higher salinities, fringing and restored marsh had similar emergence (< 150 seedlings/m²). Results indicate that recruitment is likely to be more successful under drawdown conditions, and in low salinity conditions. However, both a lack of SAV emergence in the germination experiment and a lack of SAV seeds in the seed banks using the sieving method suggest that reliance on seed banks for the restoration of shallow water areas in southwest Louisiana may prove unsuccessful.