How do you make a choice among various options? In classical decision theory, choices among options are guided by a principle of value maximization (VM): each option χ, is assigned a value, v(χ), such that the decision maker selects the option with the highest value given an offered set of attributes or dimensions. The advantages of value-based choice are refined analysis and comparison easily. The main shortcoming of value-based choice is not appropriate to apply to the real world. Thus, researchers had designed various decision mechanisms with ecological validity from different perspectives. By summarizing these studies, we found the mechanisms underlying a choice that suffers an offered loss remain largely undefined. To address this gap in understanding, we propse and demonstrate a so-called “worth-based choice” approach to explain and predict choices that result in an apparent loss. In the present study, we define v(χ) as the “value” that is assigned by the decision maker on the offered set of attributes or dimensions. More importantly, we define χc as a set of attributes or dimensions that is not offered in each option but produced afterward by the decision maker. We also define w(χc) as the delayed “worth”, which is assigned by the decision maker on the set of attributes or dimensions that is not offered but again produced afterward by the decision maker. The decision maker selects the option with the highest “worth” rather than the highest “value”. The aim of this study is to examine whether worth-based choice exists, whether the “worth” can constitute an explanation for the possible violation of the principle of VM and its underlined mechanism. Results of three studies from behavioral, cognition and neural mechanism perspectives confirmed our hypothesis. In Study 1, we first compared the decisions of children with those of their parents to explore who practiced the principle of value maximization. Then, we tried to answer the question what makes adults choose the option with the smaller value by using Query Theory. Next, we used equate-to-differentiate model to explore how individual made a final choice. And then, we seek to examine whether the worth-based choice can account for the choice that suffers an offered loss if the underling dimension is extended to more complex social contexts. At the end of study 1, we developed a worth-proneness scale to examine ecological validity in a field study. In study 2, based on the equate-to-differentiate model, we seek to examine the cognition mechanism of worth-based choice by using an eye movement system. In study 3, we consider the neural basis of worth-based choice.