Willingness to pay (WTP) is known to be lower for remanufactured products than for comparable new products. Normative work to date has assumed that a consumer's WTP for a remanufactured product is a fraction, called discount factor, of the consumer's WTP for a corresponding new product, and that this discount factor is constant across consumers. Recent empirical research demonstrates, however, that the discount factor is not constant across consumers. This discovery has led researchers to call for an exploration of more refined utility models that incorporate heterogeneous risk preferences through elements such as risk aversion, loss aversion, and ambiguity aversion. To address this call, this article assesses each of these risk preference elements by empirically deriving WTP distributions from two interlinked studies. To provide triangulation in both the empirical method and sample, the interlinked studies employ an online survey and a laboratory experiment that elicits WTP for framed lotteries that proxy the situation of buying remanufactured products. The empirical results and robustness verifications demonstrate that a parsimonious standard utility model incorporating only risk aversion explains the WTP data reasonably well.