ABSTRACTWetland restoration projects across the United States commonly lack the species and hydrology of the targeted community being restored. A better understanding of natural wetlands would help restoration project designers achieve community composition and hydrologic regimes that more closely resemble natural conditions. We investigated two different methods to identify patterns in hydrologic regime useful to restoration design. The first method was based on success criteria commonly used by North Carolina wetland mitigation projects. The second method was based on The Nature Conservancy's Indicators of Hydrologic Alteration. To test these two methods, we collected hydrologic and compositional data across the natural variation of nonriverine wet hardwood forest stands, a rare wetland community type commonly restored in North Carolina. Our results show that hydrologic parameters from the first method, based on current success criteria used in North Carolina, are not related to community composition. Hydrologic parameters from the second method, e.g. the maximum water table level over a 3‐day span, explained much of the variation in nonriverine wet hardwood forest community composition. Our results found that clear relationships do exist between hydrologic regime and community composition. Hydrologic parameters that are related to community composition need to be used as success criteria in future restoration designs. This will ensure that projects establish the appropriate hydrologic regime necessary to foster the desired wetland community type. Copyright © 2011 John Wiley & Sons, Ltd.