The traditional soil potassium (K) testing methods fail to accurately predict K requirement by plants. The Diffusive Gradients in Thin-films (DGT) method is promising, but the relationship between the DGT-measured K pool and plant available K is not clear. Wheat (Triticum aestivum L., cv. Frame) was grown in 9 Australian broad acre agricultural soils in a glasshouse trial until the end of tillering growth stage (GS30) with different plant K demands generated by varying plant numbers and pot sizes. Different K concentrations in soils were varied by 4 rates of K fertilizer application. The relative dry matter and K uptake were plotted against the soil K test value (CaCl2, Colwell and NH4OAc and DGT K measurements). To obtain 90% of maximum relative dry matter at low root density (closest to field conditions), the critical value of the NH4OAc K method was 91 (R2 = 0.56) mg kg-1. The DGT K method was not able to accurately predict relative dry matter or K uptake due to a weak extraction force for K from soils with high CEC values. Further endeavor on increasing K extraction force of the DGT method is warranted to obtain accurate plant available K results.