The physical habitat of Juncus roemerianus was examined at nine sites along a salinity gradient in the Cape Fear River Estuary, North Carolina. Soil salinity, drainage, redox potential, pH, etevati(m, percent sand, percent organic matter, and above-ground plant biomass and height were measured at each site, and from these data, the habitat of Juncus roemerianus was determined. Atl parameters varied over the salinity gradient, with soils at upriver sites having a high sand fraction, low organic fraction, and highest redox potentials. Downriver. well-established marshes had tow sand fractions, high organic fractions, and lowest redox potentials. Canonical Discriminant Analysis indicated that each site was statistically different from other sites due to salinity, elevation, and percent organic matter.Mean standing live bion~ss was 688 g m .z. and, despite differences in physical and chemical factors among sites, biomass of Juncus roemerianus did not vary. Juncus roemerianus was found to grow equally well within a broad range of physical and chemical habitats but did not occupy the total expanse of its potential habitat at any one site.Extensive overlap in physical habitat occurred between Juncus-dominated communities and adjacent communities dominated by other species, especially in the more established marshes. However, Canonical Discriminant Analysis statistically separated short and tall form Spartina alterniflora, Distichlis spicata, Scirpus robustus, and Juncus roemerianus mJcrohabitats based on elevation and redox potential. Thus, we found zonation in tidal marshes of the Cape Fear River Estuary was based on abiotie factors, but we recognize the importance of plant species interactions and marsh position within the landscape.