Parametric distributions applied to dichotomous choice contingent valuation data invoke assumptions about the distribution of willingness to pay that may contravene economic theory. This article develops and applies distributions that allow the shape of bid distributions to vary. Alternative distributions provide little, if any, improvement in statistical fit from commonly used distributions. While median willingness to pay is largely invariant to distribution, estimates of mean consumer surplus diverge widely. Sensitivity analysis to determine benefit measure response to distributional assumptions is essential to prevent erroneous policy advice from applied dichotomous choice research.