A rationalistic approach to decision-making requires greater resources than decision-makers command. The incremental strategy, which takes into account the limited capacity of actors, fosters decisions which neglect basic societal innovations. Mixed-scanning reduces the unrealistic aspects of rationalism by limiting the details required in fundamental decisions and helps to overcome the conservative slant of incrementalism by exploring longer-run alternatives. (Incremental decisions tend to imply fundamental ones, anyway.) The mixed-scanning model makes this dualism explicit by combining (a) high-order, fundamental policy-making processes which set basic directions and (b) incremental ones which prepare for fundamental decisions and work them out after they have been reached. Mixed-scanning has two further advantages over incrementalism: It provides a strategy for evaluation and it does not include hidden structural assumptions. The flexibility of the different scanning levels makes mixed-scanning a useful strategy for decision-making in environments of varying stability and by actors with varying control and consensus-building capacities.