Seagrass cover has declined in many areas of the world in a trend that has accelerated over the past several decades. This raises concern for both the impact the decline in cover has on coastal ecosystems and the effect it may have on seagrass evolutionary potential, as genotypic and genomic variation is lost. We used 8 microsatellite loci to investigate genetic diversity, structure, and connectivity in the seagrass Halodule wrightii from the Gulf of Mexico (Texas, USA) and western Atlantic (Bermuda). We examined how estimates correlated with changes in H. wrightii abundance and distribution on the Texas Gulf coast over the past 50 yr. Results show that, compared to other species, H. wrightii from this region exhibits variable clonal diversity (R = 0.02-0.81), moderate allelic diversity (mean AR = 4.09), and relatively high heterozygosity (mean He = 0.56). The patterns of genetic diversity and structure, however, do not entirely coincide with either geography or recent historical trends in seagrass distribution in this region. Results from a basin in which seagrasses have recently been expanding were consistent with expectations, as they were for an isolated site near the limit of H. wrightii’s range. Results from basins in which seagrasses have been experiencing decline and/or fragmentation, however, were mixed. Genetic structure on the Texas coast was relatively weak and coincided more strongly with tidal range than with geographic barriers or distance. Rapid expansion and the discovery of identical multi-locus genotypes at several sites raises the possibility of migration via drifting vegetative fragments, as the geographic distance among certain multi-locus genotypes cannot be explained by rhizome growth models.