This paper offers a model for the bias found in willingness-to-pay valuations against new treatments. For example, this bias provides an explanation for patient preferences that make it difficult for formularies to take treatments off their lists, even when newer treatments would appear to be clearly preferable. The appeal of the model, which is based on imperfect information, is that it is consistent with rational preferences and rational behavior by patients, which are necessary for standard models and methods related to decision theory, cost-effectiveness, and efficiency.