The EU ETS is the world’s largest carbon market and has been operating for more than eight years. We take stock of the empirical literature examining its overall efficiency. A significant strand of this literature focuses on observable fundamental drivers for allowance price fluctuations. Although the various models differ in content and methodology, a consensus has emerged that allowance prices are significantly related to fuel prices and to variables affecting the expected amount of necessary abatement, such as economic activity or changes in the cap. However, the relationship is not robust, probably because the relevant abatement technologies change with the economic conditions they operate in. Models explicitly accounting for uncertainty about future demand and supply of abatement seem better in explaining allowance price variation during certain periods. Yet, our understanding of the level of the allowance price remains poor. We cannot say with any degree of confidence whether the price is “right,” in the sense that it reflects marginal abatement costs, or whether there is a price wedge caused by transaction costs, price manipulation, or other sources of inefficiency. It would also be difficult to predict the price effects of non-marginal changes such as adjustments to future emissions caps or of overlapping policies that affect CO₂ emissions. Nevertheless, the market has matured compared to Phase I, and the banking provision has induced it to incorporate future scarcity of allowances and to smooth the effect of transient shocks as intended.