During the last two decades, the general superiority of Expected Shortfall (ES) over Value-at-Risk (VaR) has been highlighted frequently. Nevertheless, VaR remained the fundamental regime for regulatory internal model approaches to market risk under Basel I, Basel II(.5) and Basel III for more than 15 years. Just recently, the Basel Committee on Banking Supervision finally announced to move to ES taking into account the experiences made during the financial crisis and the significant weaknesses of VaR. Their resistance can be attributed to the specific characteristic of ES as it cannot be easily backtested compared with VaR. While a lot of different backtesting methodologies have already been identified for VaR, less work has been done for ES. In 2011, it was even shown that ES does not fulfill the requirements of an elicitable risk measure, which was interpreted as the fundamental proof that ES could not be backtested. This thesis presents and evaluates three alternative backtesting approaches which could even be used for backtesting ES. The thesis finds clear evidence against the criticism of ES by showing that backtesting is generally possible and that concrete tests can be applied to evaluate market risk model performance. The results will also offer added-value to regulators as first thoughts will be provided on the question of how this test could be converted into a traffic light approach similar to what is currently in place for regulatory VaR backtesting.