Legislation mandating the protection of wetlands and current pressures to convert them to other uses emphasize the need to accurately determine a wetland-upland boundary. The authors investigated six methods designed to establish such a boundary based on vegetation. Each method was applied to a common data set obtained from 295 quadrats along 22 transects between marsh and upland in 13 Oregon and Washington intertidal saline wetlands. The multiple occurrence, joint occurrence, and five percent methods required plant species to be classified as salt marsh, upland, and non-indicator; cluster and similarity methods required no initial classification. Close agreement between wetland-upland boundaries determined by the six methods suggests that preclassification of plants and collection of plant cover data may not be necessary to determine a defensible boundary. Examples of each method and lists of indicator plant species for coastal California, Oregon, and Washington are provided.