Carolina Bays are shallow depression wetlands found in the southeast US that have been severely altered by human activity. The need to restore these complex and diverse systems is well established, but our understanding of basic wetland hydrological processes is limited, hence our ability to predict the need for and/or assess the effectiveness of bay restorations is hindered. Differing physicochemical properties of soils within bay interiors may control bay hydrology. However, previous efforts to establish relationships between soil characteristics and bay hydrology have been inconclusive and the question still remains as to why some bays are ponded throughout the year while others, within a similar landscape unit, are predominantly dry. An assessment of soil and hydrologic characteristics was initiated in restored and unrestored control bays to determine if a relationship exists. Soil morphology was described and permanent monitoring wells were installed at each site. Soil samples were collected by horizon to a depth of 2 meters at the topographic center of each site, and then analyzed. After three years, multiple regression analysis (stepwise backward and forward) was used to establish relationships between the soil physicochemical characteristics and bay hydroperiod in the undisturbed sites. Results from surface soils indicated that exchangeable acidity (EA) was the best single predictor of hydrology. The best double predictor was EA and total N and EA, total N and total C as the best triple predictor. A significant relationship (r2 = 0.96) between hydroperiod and clay content in the argillic horizon (Bt) was also observed. Subsequently, this relationship was utilized to predict hydrologic response using pre-restoration hydroperiod data. The model accurately identified sites that did not need hydrologic restoration (too wet), and effectively showed sites that responded well to restoration activities.