Recent field trials on alkaline soils in southern Australia showed significant grain yield responses to liquid compared with traditional granular forms of P fertiliser. However the advantages of liquid over granular P forms of fertiliser has not been consistent on all soil types. In order to better predict the soil types on which liquid P fertilisers are likely to have potential, a glasshouse trial was conducted to compare the responsiveness of wheat to both liquid and granular forms of P on a wide range of Australian soils. A granular P fertiliser (triple superphosphate) and 2 liquid fertilisers (phosphoric acid and ammonium polyphosphate) were compared at a rate equivalent to 12 kg P/ha in 29 soils representing many of the soil types used for grain production in Victoria and South Australia. Wheat biomass was enhanced by P application in 86% of the soils tested. In 62% of the P-responsive soils, wheat dry matter was significantly greater when liquid P fertilisers were used compared with the granular form. Chemical analysis of the soils tested showed that the better performance of liquid P forms was not correlated to total P concentration in soil, P buffer capacity, or P availability as measured by Colwell-P. However, there was a significant positive relationship between calcium carbonate (CaCO3) content of soil and wheat responsiveness to liquid P fertiliser.