In a context where several sectors of society compete for space, land use types must be carefully designed and spatially allocated to guarantee a sufficient level of relevant ecosystem services (ES) in a territory of interest. In this respect, contemporary land use planning involves multiple, often conflicting objectives and criteria. Consequently, major benefits can be expected from spatial decision support systems (sDSS) designed to deal with complex spatial allocation problems. This paper presents the generic conceptual framework 'OSMOSE' and its free and open source software implementation, for the generation of specific sDSSs for spatial land use planning. The specific sDSSs generated with OSMOSE are meant to (i) identify land units which meet multiple predefined ES-attribute values for a specific land use type (LUT) and (ii) rank land units for a given LUT according to these multiple ES-attributes. A complementary purpose is to (i) identify and (ii) rank LUTs for a given land unit. Whereas 'identification' is done by means of a threshold query, 'ranking' is based on the Iterative Ideal Point Thresholding (IIPT) method. The proposed framework is extremely flexible as it can accommodate differentially weighted, continuous and/or ordinal attributes with, for the latter, equal or unequal number of classes, alternative land unit definitions and land use types. Moreover, assessments cannot only be made using ES-levels for the land unit/LUT combinations but also in terms of changes in ES-levels after a particular change of LUT. The OSMOSE framework is illustrated by means of the specific sDSS BoLa which is generated to sup-port land use planning in the region of Flanders (Belgium) with focus on soil protection. Four cases are presented in which the decision support varies between the available approaches (threshold – selection, threshold – ranking, IIPT – selection, IIPT – ranking).