The economy is more and more frequently regarded as a complex system of interacting agents. Recent developments of this approach have focused on three main issues: i) The heterogeneity of agents in the economy: The representative agent model, while elegant, versatile and widely accepted, is unrealistic and does not capture the variety in economic behaviour. ii) The ways in which agents interact: Powerful results have been derived for market-mediated economic interactions or for strategic interactions in game-theoretic settings with few agents. Many socio-economic issues call for intermediate approaches where economic interactions are non-market, non-strategic and distributed (in space-time). Actually, the network of socio-economic interaction and its dynamics has become a subject of interest in its own. iii) The dynamic process which governs the evolution of the individual: The classical model where individual behaviour arises as the optimal contingent plan of actions of a deductive rational, perfectly informed utility maximizer agent, is more and more frequently replaced by models where agents learn and adapt to their economic environment. This raises the issue of understanding the collective dynamical properties of systems of boundedly (inductively) rational interacting agents. This Workshop offers a forum for presentation and discussion of the latest results on these issues. KEYNOTE SPEAKERS: J.-P. BOUCHAUD (CEA-Saclay & Science and Finance, Paris, France) S. METCALFE (University of Manchester, U.K.) J. SCHEINKMAN (Princeton University, U.S.A.) H.E. STANLEY (Boston University, U.S.A.) Y.-C. ZHANG (University of Fribourg, Switzerland)