The application of the market microstructure theory to foreign exchange markets in the last few years has introduced a new approach to the analysis of exchange rates. The most important variable of the microstructure analysis, the so-called order flow has proven to be suitable for explaining a significant part of exchange rate changes, not only for high frequency data, but also at longer time horizons that are relevant for macro-economic analysis. Microstructure theory is thus extremely successful from an empirical point of view, especially when compared to traditional exchange rate models. The aim of our study is to provide an introduction to the microstructure-based analysis of exchange rates, emphasising those aspects which may be the most relevant for central banks. In addition to an introduction to the theoretical background of the microstructure approach and the presentation of the key empirical results, we also intend to cast light upon the questions which are important for central banks and which can be tackled successfully using this framework. On the basis of the literature's findings, we present the answers given by the microstructure approach to, among others, questions concerning the efficiency of central bank intervention, the effects of economic news on exchange rates, and the role of different currency market participants in exchange rate developments.