Most soils in eastern Australian contain abundant soil potassium (K) reserves, and it is often assumed that there are no problems with soil K status. However, soil K deficiency has been reported in selected locations, and for viticulture, there are potential problems with high soil K concentrations due to the application of winery wastewater. This study investigated different soil K variables and plant variables for six soils with distinctive properties from across central and southern New South Wales to determine the presence of soil K deficiency and to understand the effect of adding K on the dynamics of soil K availability. A glasshouse experiment compared the selected soils under three fertiliser K rates with forage kale as the test species. Highly significant differences (P < 0.001) were found for soil and fertiliser K rate effects for three measures of soil K (solution K, soln K; exchangeable K, exch. K; tetraphenyl borate K; TBK). Significant soil and fertiliser rate effects were detected (P < 0.001 and P = 0.04 respectively) for the plant shoot (stem and leaf) biomass and nutrient uptake efficiency (UPE) index, but no plant K deficiency was detected; in fact, luxury K consumption was likely. Quantification of K efficiency indices (UPE and utilisation efficiency, UTE) demonstrated significant differences between the soils in the ease with which K was removed. This was illustrated by the negative correlation between both UPE and UTE with final exch. K. From soil properties potentially related to soil K variables, a significant linear regression relationship (P = 0.05) was found for TBK with illite and clay content. By contrast, a linear regression relationship between exch. K and illite content only was weaker (P = 0.09). These relationships show how soil properties (especially mineralogy) can predict soil K variables. A significant positive log-log relationship was found between exch. K or TBK for 37 Queensland soils and the same soil K measures from this study, consistent with this relationship. This relationship indicates that TBK can be effectively predicted from measuring exch. K for a wide range of soils across eastern Australia, but more research is required to understand the value of TBK to predict soil K availability.