We present a real options model of an agent's decision to purchase or sell a good under conditions of uncertainty, irreversibility, and learning over time. Her WTP and WTA contain both the intrinsic value of the good and an option value associated with delaying the decision until more information is available. Consequently, the standard Hicksian equivalence between WTP/WTA and compen-sating and equivalent variation no longer holds. This helps to explain the WTP/WTA disparity often observed in laboratory experiments and surveys because subjects may have limited learning time and opportunities, thus generating option values. In contrast, the disparity may decrease or disappear entirely in real markets since agents are free to choose when to stop gathering information, thereby reducing or eliminating the option values. We also discuss experimental evidence that supports the option value hypothesis.