A variety of questioning formats have been used in contingent-valuation studies, with dichotomous-choice questions becoming the preferred format. However, as with any empirical technique, continued applications raise questions that require attention if the credibility of the procedure is to be maintained. It is shown that estimated Hicksian surplus can be substantially affected by the selection of a functional form when analyzing responses to dichotomous-choice questions. Given that theory, intuition, and empiricism all play a role in developing these estimates, several maxims are suggested for evaluating and/or mitigating such effects in future studies.