This paper investigates how uncertainties related to natural gas prices and potential climate policies may influence capacity investments, utilization, and emissions in US electricity markets. Using a two-stage stochastic programming approach, model results suggest that climate policies are stronger drivers of greenhouse gas emission trajectories than new natural gas supplies. The dynamics of learning and irreversibility may give rise to an investment climate where strategic delay is optimal. Hedging strategies are shown to be sensitive to the specification of probability distributions for climate policy and natural gas prices, highlighting the important role of uncertainty quantification in future research. The paper also illustrates how this stochastic modeling framework could be used to quantify the value of limiting methane emissions from natural gas production. (C) 2014 Elsevier Ltd. All rights reserved.