The article discusses how economic agents learn to have rational expectations. Instead of needing perfect initial information, they can learn within certain limits. Methods from adaptive control theory help simplify the analysis. Different starting points and learning methods can lead to various outcomes in economic models.