Traditional theories of cognitive development predict that children progress from intuitive to computational thinking, whereas fuzzy-trace theory makes the opposite prediction. To evaluate these alternatives, framing problems were administered to preschoolers, second graders, and fifth graders. Consistent with fuzzy-trace theory, results indicated (a) that younger children focused on quantitative differences between outcomes and did not exhibit framing effects (risk avoidance for gains, risk seeking for losses) and (b) that older children assimilated these quantitative differences and displayed framing effects.