Many criticisms have been made of optimization theory (Laville, 1999a). These objections may be explained by the fact that human rationality is bounded – that decisions are constrained by cognitive limitations (Simon, 1982). In the present paper, I will show that if rationality is bounded, then we must study the processes of decision. My thesis is that cognitive limitations lead to procedural rationality. Although this assertion has already been sustained implicitly by Simon (1959) and explicitly by Mongin (1986), it has not been argued in sufficient depth.