This study suggests a model for calculating total benefit rigorously to use the contingent valuation method (CVM) in cost-benefit analysis (CBA). Estimating households’ willingness to pay through survey method, the study attempts to demonstrate if a respondent’s income and the distance between a respondent’s residence and the location of a target facility affect her willingness to pay. The estimation results from a structural model show that income and distancedecay effects exist and that the calculated total benefit varies largely when these effects are ignored. The study emphasizes the effects of income and distance-decay on the total benefit must be carefully considered in using CVM for CBA. Even though the total project cost is precisely estimated, the benefit/cost (B/C) ratio may differ largely when the total benefit is not correctly calculated. Also, an ad hoc model generates significantly different estimates from the utility difference model this study adopted. The difference in estimates suggests that the total benefit has to be estimated by a structural model. Finally, simulations are performed to check the validity of the model as well as to predict consequences when income and distance-decay effects are not properly treated. The results from simulations reveal it is not desirable to ignore those effects considering the perspectives of balanced regional development.