This book analyses the impact of learning behaviour in different economic environments. All cases are motivated by real world puzzles which cannot be explained by models with rational expectations: The size and persistence of exchange rate volatility, the possibility of financial market crashes without news, and progress in disinflation in transition countries despite limited credibility and conflicting monetary and fiscal policies. In particular, the author shows that adaptive learning behaviour adds additional dynamics to economic models, which are not present under the assumption of full rationality. Certain types of hysteresis in financial markets can be explained by relaxing the rational expectations assumption; credibility and transparency of economic policies matter even more when agents learn about their economic environment; and markets with incomplete and asymmetric information are inherently unstable and prone for market crashes.