Multi-criteria decision making techniques are often used in the field of water resources. Their function is to facilitate decision making for the purpose of selecting the best solution to a particular problem from a set of potential alternatives. ln order to assist in the selection, multicriteria decision making techniques evaluate each of the potential alternatives. The evaluation is based on an assessment of how well each of the alternatives satisfies specified criteria. These criteria typically are the characteristics of the alternatives, or consequences which would occur due to implemententation of the potential alternatives. Often the measures of the criteria, or criteria values, associated with the alternatives have an uneven spatial distribution. For example, implementation of a particular alternative could produce favorable impacts in one location in a given region, while resulting in negative consequences for other areas. As a result, the best alternative for one area within the given region may not be the best solution for all locations in that region. In the evaluation of alternatives by conventional multi-criteria decision making techniques this spatial variability in the criteria values is often not taken into consideration. The criteria values used by conventional techniques typically represent the average characteristics of the alternatives, or total impacts produced by the alternatives for the entire region. Thus, in evaluating potential alternatives, the localized characteristics and impacts associated with the alternatives are not taken into consideration. As a result, the alternative selected as best using the multi-criteria decision making techniques may have significant negative characteristics or impacts in specific areas within the region. This shortcoming in conventional multi-criteria decision making techniques is demonstrated in this study using a floodplain analysis of the Red River Valley near the City of Winnipeg, Manitoba, Canada. ln this study a set of potential flood protection alternatives are generated for a region within the Red River Valley. Each of the potential alternatives in the set are evaluated and ranked on the basis of multiple criteria. The criteria used in this evaluation are impacts to the region produced by flooding which would occur with implementation of each of the various alternatives. The evaluation of the alternatives is conducted using two multi-criteria decision making techniques. First, the alternatives are evaluated and ranked using the Compromise Programming technique. In this evaluation the spatial variation in the criteria values associated with the alternatives is not considered. The second multi-criteria decision making technique used in this evaluation was the Spatial Compromise Programming technique. This new technique was developed as part of this research by combining GIS technology with the Compromise Programming technique. Using the Spatial Compromise Programming technique it was possible to account for spatial variability in the criteria values used in the evaluation of the potential flood protection alternatives. By comparing the results of the two multi-criteria decision making techniques it is shown that the spatial variation in the criteria values must be taken into consideration in order to provide an accurate evaluation of the potential alternatives.