In the Okavango Delta 98-99% of the water from inflow and rainfall is lost to the atmosphere through evapotranspiration. As a consequence 94% of inflowing solutes are retained within the Delta landscape. This process might be expected to result in an entirely saline environment, but that is not the case: the surface waters have very low salinity, supporting a typical freshwater biota. It has been deduced that the numerous islands in the Delta (about 150,000 within an area of 13,500 km2) have been formed through evapotransporative concentration in the groundwater, of infiltrating solutes, followed by precipitation and volume increase. Evidence of this is the large amount of calcrete in island soils. These islands of 3-10 m thickness with clayey soils are underlain by fine Kalahari sand to a depth of 200-300 m, which also indicates that they are formed through surface processes. The infiltration rate of surface water from floodplains and streams into islands is very high, and is predominantly a lateral process that is unidirectional. Evapotranspiration in the riparian woodland zone cause the ground-waters in the central area of islands—with halophyte grasslands—to have very high salinities. By use of chloride as a conservative element the concentration factor between central island groundwater and surface water is calculated to be 500-1,000. This groundwater is depleted of calcium and magnesium supporting the early deductions that these elements have precipitated as calcrete. There is also a large depletion of silicate and potassium that probably have precipitated as well forming the clayey soils typical of the islands. The central island groundwater is dominated by sodium, bicarbonate and dissolved organic matter. The gradual increase of salinity here causes a periodic let off of this water through a density-driven process to deeper layers. This process together with island growth through precipitation of solutes are the two major sink processes of inflowing solutes and explains why the Okavango Delta is at present a freshwater system. The whole island complex is calculated to be 100,000-400,000 years old while some intensely studied islands may be younger: 80,000-240,000 years. The discrepancy is explained by a biassed selection of islands currently in flooded areas with better growth conditions. The uniqueness of the Okavango Delta and ideas for future research are discussed.