The decreasing mineral concentrations in the grains of cereals have recently stimulated research to better understand the cropping determinants of grain mineral composition. This study aimed to analyze the effects of liming on the mineral concentrations in the grains of three cereal crops: barley, oat, wheat. The hypothesis tested was that soil pH is the main driver of the grain nutrient concentrations in crops, through its influence on the soil extractable minerals. Macro nutrients (Ca, K, Mg, P, S), micro-nutrients (Cu, Fe, Mn, Se, Zn) and some trace elements (As, Cd, Pb) were analyzed. Two long term liming trials in SE England (1962 -) were studied, with the same crops sown in the same years. On each site, four liming rates were applied to 32 plots to create a pH range from approximately 4.5 to 7.5. The trials were subdivided into two P fertiliser treatments, consisting of a nil and regular P inputs. For a given crop, the effects of pH, soil type, concentrations of nutrients in soil extracts and of P treatment on the grain mineral concentrations were tested. This pairwise analysis was followed by a multiple linear regression analysis in order to determine the main explanatory variable for crop mineral concentration. Liming had a significant impact on most of the soil extractable mineral concentrations, except extractable K and Mg. The grain mineral concentrations exhibited significant differences between crops, the concentrations in wheat being the smallest. pH proved to have a larger direct effect on mineral concentrations in grain (e.g. Ca, Mg, P, Mn) than through its influence on extractable nutrients (e.g. Cd). Grain nutrients responses to pH were, however, not the same in the three crops. Differences in Cu and Zn were mostly accounted for by the effect of soil type, the soil with the higher CEC leading to the higher grain concentrations. For Fe, Pb and K, no correlation could be found between the grain mineral concentrations and the explanatory variables. Difficulties in explaining the grain mineral concentrations are due to specific crop responses to nutrients, usefulness of soil extractions, and complex physiological processes in mineral translocation from roots to grains. The results underline the difficulty of using ordinary soil analysis for predicting the quality of cereal grains for nutrition, and caution in the use of grain testing to recommend soil fertility enhancing practices.