There is an increased demand for quantitative high-resolution soil maps that enable within-field management. Commonly available soil maps are generally not suited for this purpose, but digital soil mapping and geophysical methods in particular allow soil information to be obtained with an unprecedented level of detail. However, it is often difficult to quantify the added value of such high-resolution soil information for agricultural management and agro-ecosystem modelling. In this study, a detailed geophysics-based soil map was compared to two commonly available general-purpose soil maps. In particular, the three maps were used as input for crop growth models to simulate leaf area index (LAI) of five crops for an area of â¼ 1 km.sup.2 . The simulated development of LAI for the five crops was evaluated using LAI obtained from multispectral satellite images. Overall, it was found that the geophysics-based soil map provided better LAI predictions than the two general-purpose soil maps in terms of correlation coefficient R.sup.2, model efficiency (ME), and root mean square error (RMSE). Improved performance was most apparent in the case of prolonged periods of drought and was strongly related to the combination of soil characteristics and crop type.