By 2050, the world's population will have reached 9 billion. To feed that many people, soil fertility will have to be maintained artificially. All fertiliser materials depend on a geological resource: nitrogen (N) fertilizer production needs fossil fuels, and both phosphate (P) and potassium (K) are derived by mining. Irrespective of new biological techniques in plant breeding and genetic modification, soils still need to supply the mineral nutrients that plants require, and these are exported from soil with every harvest.Studies of global offtake of N, P and K from soils through crop production show that although N and P are roughly in balance, removal of K from soils greatly exceeds inputs. World mine production of K needs to double to replace the amount removed in crops. Recent revision of reserve estimates for potash and phosphate rock show significant increases for phosphate rock and reductions for potash. Potash supply is now potentially of much greater concern than phosphate.Against this background, it is clear that new potash mining ventures are required. In the developed world, the supply of potash from conventional sources will continue. However, in other countries the high price of potash means that novel unconventional sources are being considered. K silicate minerals (such as micas, feldspar and nepheline) have the potential to provide an adequate source of K for communities that cannot afford conventional fertiliser. However, it is not the total K content of these materials that controls their ability to supply plant nutrients, but the rate at which they dissolve.