Four soils, varying in calcium and magnesium content, were treated with KH2PO4 or K2HPO4 pellets. Crystalline phosphate reaction products were identified by means of X-ray diffraction analysis over a period of 450 days. The addition of KH2PO4 resulted in the formation of DCPD in all four soils. However, DCPD formed only in two soils treated with K2HPO4. DMPT formed along with DCPD in soils having a water-soluble calcium to magnesium ratio of less than 1.5. OCP was a prominent reaction product in all soils. A calcium-magnesium phosphate, Ca3Mg3(PO4)4, previously not reported as occurring in soils, was detected in the soil containing approximately equivalent amounts of ammonium acetate-extractable calcium and magnesium. A calcium phosphate, Ca4P2O9, also previously not reported, formed in soils high in exchangeable calcium... The solubility of phosphorus in the above systems (KH2PO4 or K2HPO4 added as a pellet to each of the four soils) was determined by equilibration of the soil-fertiliser system with water. Similarly, the solubility of phosphorus was determined at intervals in each of the soils after KH2PO4, K2HPO4, DCPD or DMPT was mixed throughout the soil. The solubility of added phosphorus was near that of DCPD or DMPT when the soils were incubated for a short period. The solubility of phosphorus decreased with time and approximated that of OCP after 12 to 15 months of incubation. Added phosphorus apparently remained more soluble when KH2PO4 rather than K2HPO4 was added in the pellet form. Mixing these carriers with the soil eliminated any differences which were evident for the pellet application. It was apparent that the addition of the actual initial reaction product (DCPD or DMPT) rather than a water-soluble phosphate (KH2PO4 or K2HPO4) did not greatly affect the solubility of phosphorus in a given system, i.e. the solubility of DCPD or DMPT added to soils was similar to that of applied KH2PO4 or K2HPO4.