Hedonic valuation of quality attributes can be misleading when the assumption that these attributes are exogenous to sample selection is violated. This paper considers the simultaneity between hedonic valuation and sample selection in the context of a model of consumer behavior over packaged goods and investigates empirically how the decision on house location (urban/rural) affects the household's valuation of water-related characteristics of the house in question. The empirical analysis uses data from the Indonesian housing market and suggests that households value access to safe and improved domestic water sources. However, failing to correct for sample selection results in a biased valuation of willingness to pay for house characteristics. This might misguide policy recommendations for improved provision of domestic water, based on cost-benefit analysis.