The literature has identified three main approaches to account for the way exchange rate regimes are chosen: i) the optimal currency area theory; ii) the financial view, which highlights the consequences of international financial integration; and iii) the political view, which stresses the use of exchange rate anchors as credibility enhancers in politically challenged economies. Using de facto and de jure regime classifications, we test the empirical relevance of these approaches separately and jointly. We find overall empirical support for all of them, although the incidence of financial and political aspects varies substantially between industrial and non-industrial economies. Furthermore, we find that the link between de facto regimes and their underlying fundamentals has been surprisingly stable over the years, suggesting that the global trends often highlighted in the literature can be traced back to the evolution of their natural determinants, and that actual policies have been little influenced by the frequent twist and turns in the exchange rate regime debate.