Schooling is widely seen as critical for income generation in all types of economies. A growing concern among many has been the possibility of increasing inequality in part due to children from higher-income households obtaining more schooling and reaping greater gains from schooling than children from lower-income households. There are many empirical studies for various societies that tend to find significantly positive, but small associations between household income and schooling. But these studies generally have three major limitations for the purpose of characterizing the degree of association between household income and schooling-related investments: (1) use of income indicators that may be contaminated by relatively large measurement errors and endogeneity, (2) inclusion of other household, community and schooling variables that may represent part of the association with income in empirical estimates, and (3) use of limited indicators of schooling. This paper uses a rich new household survey-commune-school facility survey from Viet Nam to illustrate how important these limitations may be. The estimates suggest: (1) predicted income (expenditure) tends to yield estimates of much stronger associations than does current income or expenditures, (2) controlling for variables such as in most previous studies reduces the estimated associations with income substantially, and (3) including a wide range of schooling-related variables leads to more nuanced understanding of income-schooling associations, with some benefits for children from poorer households but a dominant tendency for school and private behaviors to favor significantly and in many cases substantially children from higher-income households.