This chapter explores the relative merits of a number of alternate approaches to estimating Value at Risk (VaR) for electricity markets. The distinctive features of electricity markets present non-trivial challenges for the trading and hedging activities of market participants. Compared to traditional approaches to forecasting VaR, the empirical findings provide strong support for the use of Extreme Value Theory (EVT). However, more sophisticated conditional EVT approaches do not necessarily outperform vanilla EVT approaches. Furthermore, the left tail of the return distribution proves particularly challenging to model. Given the idiosyncrasies of each electricity market, it is unlikely that a single approach is optimal across the board.