We examined historical patterns of land use of depression wetlands (Carolina bay and bay-like wetlands) to determine if a relationship between vegetative successional changes over a 41-year period and previous human disturbances (primarily agricultural) could be established. Land cover was interpreted from 1951 (black and white) and 1992 (false color infrared) aerial photography of the Savannah River Site (a 780 km 2 federal nuclear facility in which wetlands have been relatively undisturbed since 1951). Patterns of change from one land cover to another were detected by constructing a series of frequency tables. About one fourth of the 299 wetlands identified were either pasture or cultivated in 1951, and the majority had been ditched for drainage. Agriculturally disturbed wetlands primarily became mixed hardwood/pine or were converted to pine plantations by 1992; however, no successional differences between wetlands that were cultivated versus pasture were detectable. The type of land use of many of the depression wetlands prior to 1951 probably was determined by physical characteristics of the wetlands (e.g., topographic position, size, and hydrologic features). Thus, in many cases, separation of recovery trajectories from other successional pathways, initial hydrogeomorphic differences, and/or continued human influences is not possible in this study. However, from this change-detection study, we recognize that many of the currently protected depression wetlands at SRS were disturbed by agricultural practices or were impacted by hydrologic alterations prior to t951, implying considerable resilience in the recovery toward a functioning wetland condition if hydrologic regimes are restored. A significant finding of this study is the relative stability of herb-dominated bays, which indicates that this vegetation type is not necessarily a successional continuum toward an eventual hardwood forest, at least in the temporal scale of the study. Thus, we suggest that management prescriptions for the restoration/conservation of herb-dominated wetlands should incorporate concepts of temporal stability within a framework of cyclical hydrologic and vegetation changes.