This study aims at developing an operational approach-namely, directional spectral mixture analysis (DISMA)-for retrieving vegetation parameters like fractional vegetation cover (FVC) and leaf area index (LAI) from multispectral and multiangular data. The approach attempts to highlight the consistency of one-dimensional models and linear mixture approaches. DISMA combines spectral signatures of soil and vegetation components with an analytical approximation of the radiative transfer equation, giving rise to a fast invertible bidirectional reflectance distribution function (BRDF) model of discontinuous canopies. Both the forward model and its inversion using a simple technique based on lookup tables method are tested using airborne POLDER and HyMap data corresponding to cropland. The method has proven fast enough to image vegetation properties over large areas, providing accurate and stable maps of FVC and LAI. The retrievals of LAI correspond well with ground-based measurements of specific crops, with root mean square error differences of 0.5-0.6 and a r2 of the linear fitting around 0.92. Though the accuracy assessment of retrieved parameters may be sensitive to the BRDF sampling, the results of model inversion remain consistent when varying the angular and spectral configurations. A model intercomparison exercise has been carried out using different models, either purely descriptive or based on radiative transfer modeling, indicating that this general approach is both sound and consistent. Thus, DISMA appears to be a useful tool for exploiting the synergistic spectral and angular potential of the new sensors.