Abstract : This article discusses recent advances in agent-based modelling applied to
macroeconomic analysis. I first introduce the building blocks of agent-based
models. Furthermore, by relying on examples taken from recent works, I argue
that that agent-based models may provide complementary or new lights with
respect to more standard models on key macroeconomic issues like endogenous business cycles, the interactions between business cycles and long-run
growth, and the role of price vs. quantity adjustments in the return to full
employment. Finally, I discuss some limits of agent-based models and how they
are currently addressed in the literature.