The empirical relation between happiness and religiosity is considered from the perspective of basic utility theory. An unbalanced cross-country panel data set is used to study whether religiosity can be considered as a substitute in the happiness function, which itself is held to be a proxy for the utility function. We find that the same level of happiness can be maintained with high and low levels of religiosity as a result of substitution along a standard indifference curve. Our empirical results are consistent with three stylized facts of the empirical literature, namely a positive correlation between happiness and religiosity, a positive correlation between happiness and income, and a negative correlation between religiosity and income.