The often observed empirical divergence between WTA and WTP measures of welfare change continues to be a topic of interest to both theoretical and applied economists. The divergence has particularly important implications for environmental policy. In this article, we review behavioral and other explanations of the disparity, with a focus on their connections to neoclassical welfare theory, and evaluate the empirical evidence of these explanations through the same lens. Some explanations of the disparity are consistent with neoclassical models, and some are not. Likewise, some imply that the divergences are attributed to underlying preferences (neoclassical or not), whereas others suggest that the divergences are due to elicitation methods, cognitive limitations, or other non-preference-related reasons. We argue that the source of the divergence can inform the choice of which measure, WTP or WTA, to use in a given empirical application.