In 1999, the Chicago Mercantile Exchange (CME) created a weather derivative market to enable businesses to transfer risk that could be adversely affected by unanticipated temperature swings. The temperature-related weather derivative market is now ever expanding with derivatives actively traded on the CME for 44 cities worldwide. It is now well-known that standard approaches to arbitrage-free pricing (e.g. Black-Scholes) cannot be applied in the weather derivative contexts. Many have thus successfully argued that the only reliable way to price weather options is by using forecasts of the underlying weather variable. In this talk, we review the literature on forecasting temperature for derivatives pricing and present a comparative study. Furthermore, we cast the average temperature forecasting problem as one in the analysis of extreme values and show that extreme-value-based models can outperform other methods in some cities.