Expert elicitations are critical tools for characterizing technological uncertainty, since historical data on technical progress may not provide a sufficient basis for forecasting future advances. The objectives of this paper are to describe the protocol and results for an expert elicitation on the future performance of gas-turbine-based technologies in the electric power sector and to discuss how these insights relate to the current elicitation literature in energy modeling. Elicitation results suggest that prospective efficiency gains are likely to be slower than historical trends; however, the assessed values are still appreciably higher than the efficiencies used in many energy models. The results also indicate that conducting face-to-face elicitations may be important for minimizing overconfidence and for critically examining reported values, especially when assessing non-central probabilities in the tails of a distribution.