Total phosphorus (P), inorganic P (P-i), organic P, and several P-i fractions were determined for 8 fertilised calcareous soils under agriculture and their virgin analogues under natural bush to ascertain changes due to agricultural development. The relationships between soil properties and forms of P were also determined. In general, agricultural development of soils resulted in increases in total P (average 105% increase), P-i (154%), organic P (49%), Olsen P (200%), Colwell P (100%), and all P-i fractions compared with their virgin analogues. For the virgin soils, the abundance of the P-i fractions was in the order: Al-P>O-P (occluded P)>Fe-P>Ca-10-P = Ca-2-P>Ca-8-P, which changed to Al-P>Ca-8-P>Ca-2-P>Ca-10-P>Fe-P>O- P for fertilised soils. The average contribution of each fraction to the increase in total P-i was Al-P (29%), Ca-8-P (26%), Ca-2-P (18%), Fe-P (13%), Ca-10-P (13%), and O-P (4%). The change in Cas-P was closely correlated with the content of the active fraction of calcite in the soil (ACCE). The increase in Fe-P associated with agriculture was highly correlated with citrate-dithionite-bicarbonate (CDB) extractable Fe (Fe-d) and acid-ammonium oxalate extractable Fe (Fe-o). The increase in Al-P was correlated with the ratio of acid-ammonium oxalate extractable Al (Al,) to Fe-o. Both Olsen and Colwell NaHCO3-extractable P were highly correlated with Ca-2-, Al-, Fe-, and Ca-10-P, and total P values. Multiple regression analysis indicated that Ca-2-P and Ca-10-P were major contributors to available P as determined by the Olsen and Colwell soil tests.