AbstractAdequately evaluating the success of coastal tidal marsh restoration has lagged behind the actual practice of restoring tidally restricted salt marshes. ASpartina‐dominated valley marsh at Barn Island Wildlife Management Area, Stonington, Connecticut, was tidally restricted in 1946 and consequently converted mostly toTypha angustifolia.With the re‐introduction of tidal flooding in 1978, much of the marsh has reverted toSpartina alterniflora.Using a geographical information system (GIS), this study measures restoration success by the extent of geographical similarity between the vegetation of the restored marsh and the pre‐impounded marsh. Based on geographical comparisons among different hydrologic states, pre‐impounded (1946), impounded (1976), and restored (1988) tidal marsh restoration is a convergent process. Although salt marsh species currently dominate the restored system, the magnitude of actual agreement between the pre‐impounded vegetation and that of the restored marsh is only moderate. Further restoration of the salt marsh vegetation may be limited by continued tidal restriction, marsh surface subsidence, and reduced accretion rates. General trends of recovery are identified using a gradient approach and the geographic pattern’ of vegetation change. In the strictest sense, if restoration refers only to vegetation types that geographically replicate preexisting types, then only 28% of the marsh has been restored. Restoration in a broader sense, however, representing the original salt marsh vegetation regardless of spatial position, amounts to 63% restored. Unrestored marsh, dominated byTypha angustifoliaandPhragmites australis, remains at 37%. By emphasizing trends during vegetation recovery, this evaluation technique aims to understand the restoration process, direct future research goals, and ultimately aid in future restorati