<p>Motivated by the increasing interest of academics and practitioners in swing options, we develop a method for computing swing option-implied volatility. In a dynamic programming option pricing framework supplemented by an additive single-factor forward curve model, we propose to obtain implied volatility via a combination of Monte Carlo techniques and a root-finding algorithm. In addition to deriving the convergence properties of our approach, we apply it in an empirical study of volatilities implied by natural gas swing options. Here, we investigate their key features (related to seasonality and moneyness) and highlight their industry merits in delta hedging applications.</p>