Wetlands provide many important ecosystem functions and services worldwide and are hotspots of biological diversity. However, depressional wetlands are particularly vulnerable to effects of climate change due to the significant role that precipitation and surface runoff play in shaping their hydrology. In the Southern Great Plains of North America, climate projections predict more extreme storm events, higher temperatures, and severe droughts, which could threaten natural hydrological patterns of depressional wetlands in this region. Regional hydrological models that accurately predict water dynamics are critical for developing effective climate change adaptation strategies. We developed a model to predict wetland inundation status for depressional wetlands in the Pleistocene Sand Dunes Ecoregion of Oklahoma, USA, that evaluated effects of weather variables, wetland characteristics, and landscape-level variables. We then predicted numbers of inundated wetlands and frequency of wetland inundation under three climate change scenarios for the middle and end of the century (2036–2050 and 2084–2099, respectively). Total precipitation measured in the 2 months prior to an inundation event and average daily temperature were the most important variables predicting wetland inundation status, and land use and wetland characteristics explained relatively little variation in water dynamics. Projections of wetland inundation status indicate numbers of inundated wetlands will decrease in spring and summer by as much as 42% and 79%, respectively, by midcentury. Future inundation patterns during fall and winter were less clear but will likely be similar to current, highly variable conditions. These results suggest climate change may threaten persistence of wetlands during key seasonal periods when humans, plants, and wildlife depend on them for crucial resources and services.