Background Seawater is potentially a vast and readily accessible reserve of potassium for fertilizer production. Potassium is not at present recovered from seawater commercially, since potash can be produced easily and cheaply from the extensive deposits of certain soluble salts that exist in some countries. However, Australia has no known deposits, and there is interest here in the development of a method for extracting potassium from seawater, or more particularly from the concentrated bittern which remains after the crystallization of sodium chloride by solar evaporation. Objective An assessment was undertaken of the merits of possible lines of approach to potassium recovery, by reviewing published literature relating to all earlier processes or proposals. From this information, methods showing some economic promise were to be subjected to preliminary experimental investigation. Summary of Work Done From a study of the chemical literature, a review was prepared of possible methods for recovering potassium from bitterns. An experimental study was made of the extraction of alkali metals by a 4:1 mixture of Santophen-1 (a sterically-hindered phenol) and EHPA. This system, which is capable of operation at low pH, offered the possibility of a significant economic advantage over the use of the phenol alone, due to reduced alkali requirements. It was found that the extractant showed an optimum potassium-sodium selectivity of about 9 when the final pH was 5. Selectivity was lower at high salt concentrations. The alkali metals could be completely stripped from the extractant by nitric acid solutions of pH below 2. Magnesium was strongly extracted with the alkali metals from diluted bittern. Conclusions Using crystallization methods, complicated procedures would be required to produce a fertilizer-grade product from seawater bitterns. Economical operation could possibly be achieved if KC1-production were integrated into a comprehensive scheme for the recovery from seawater of NaC1, MgO and other saleable products. No selective extraction procedure has yet proved wholly satisfactory. However, the experimental findings have shown sufficient technical promise to inspire confidence that it might ultimately be possible to extract potassium selectively from seawater bittern at a comparative cost. Methods based on selective reagents of the dipicrylamine and phenolictypes are undoubtedly technically feasible, but they involve the consumption of alkali, and would not be economic at current market prices. It appeared that the use of di (2-ethyl hexyl) phosphoric acid (EHPA) in conjunction with a phenolic reagent might result in a worthwhile reduction in alkali requirements. Experimental results with Santophen-1-EPHA revealed several features which make the system unattractive for commercial application. In particular there were the interference of magnesium in potassium extraction, the required dilution of the bittern, and the quantity of alkali needed to establish the working pH in the presence of EHPA. In addition, relatively small losses of the costly reagents could become significant in view of the large volumes of brine to be treated. The properties of some inorganic ion-exchange materials suggest application to potash recovery with attractive possibilities of low production costs. The realisation of this potential would depend on the development of the ion-exchangers to satisfy practical criteria, such as adequate rates of exchange and selectivity in liquors of high salt content. This development work might well require considerable time and effort, but a useful preliminary assessment could be obtained from a limited experimental investigation. Recommendations It is recommended that a preliminary experimental evaluation be made of certain inorganic materials as potassium selective ion exchangers.