Recharge rates determined at diverse study sites in a shallow, unconfined aquifer differed from one another depending on the analytical method used and on each method's applicability and limitations. Total recharge was quantified with saturated-zone methods using water-table fluctuations at seven sites in North Carolina, USA and using groundwater-age dating at three of the seven sites; at two of the sites, potential recharge was quantified with an unsaturated-zone method using Darcy's law; and at five of the sites, net recharge was quantified with a stream hydrograph-separation method using streamflow-recession curves. Historical mean net recharge was 25 to 69% of the historical total recharge rates. The large disparity is attributed to groundwater losses between recharge and discharge areas, primarily by evapotranspiration and seepage to underlying aquifers. The spatial distribution of historical mean annual total recharge did not vary between landscape units, as suggested in a previous study. Similarly, total recharge did not correlate significantly with mean annual rainfall, mean annual water table depth, or the surficial soil properties of percent clay and bulk density. Total recharge did correlate significantly with the surficial soil properties of percent sand and percent silt. ?? Springer-Verlag 2007.