Modeling agricultural systems that recognize the environmental dimensions of agriculture has evolved during the two past decades. Multi-objective mathematical models encompassed the diversity of objectives inherent in agricultural activities as a result of externalities and replaced single objective models. It has been observed that recent modeling efforts at farm level combined several simulation models at a time (crop simulation, weed simulation, hydrologic model, erosion) with a multi-criteria model. Many of the studies reviewed have not thoroughly considered the policy instruments to internalize the pollution problems and some studies have not considered any policy instrument. The multicriteria techniques used range from distance based approach, utility theory, generating methods, interactive methods, fractional programming and fuzzy programming. The later method is called for to deal with the inexact information generated with geographical information systems GIS or simulation models. Most applications of coupled GIS and decision models dealt with watershed management and soil erosion. Spatial GIS/multicriteria models that involve stakeholders are considered as a form of institutional reorganization which will help change the hierarchical mode of decision making. Stakeholder involvement in decision making has brought the modeling effort to include group and multi- ple decision makers. It is expected that future models will integrate several simulation models, GIS, and several stakeholders and would be applied at regional and national levels. Thus, multicriteria modelers will have to deal with uncertain and inexact information as well as asymmetric information.