The purpose of child well-being indices is to distill large amounts of data on children in ways that can be easily communicated to and used by policy makers and the public. Several major methods for creating indices exist: (1) a standard score method that assesses differences across geographic areas, (2) a micro-data tally method that assesses the number of problems children have, and (3) an average percentage change method that assesses change from year to year (compared either to a base year or a “model” year). We review each type of index and discuss some cross-cutting issues, including the use of micro- versus macro-data, the measurement of absolute versus relative levels of well-being, and the constructs of well-being and context. We then use data from a single source, the National Survey of America’s Families, to demonstrate the usage of indices, comparing the relative performance of 13 American states. We find that states tend to be ranked similarly with regard to well-being and to the combination of well-being and context, regardless of whether one is examining the average concentration of problems per child, relative status of children across states, or change over time. We find that the states with relatively lower levels of well-being, as well as states whose children have the highest average number of problems, tend to show the greatest improvement over time in well-being. We also find that states in which children are more likely to experience multiple problems tend to be the same states in which the overall levels of well-being and the condition of children are relatively poorer. Despite the fact that the literature has referred to different types of indices commonly as “child well-being indices,” the differences both in the purpose and in the results across tabulation methods illustrate the importance of balancing the simplicity of the presentation of index results with sufficient information to enable people to understand exactly what the index is assessing. Furthermore, our results highlight the value of using multiple types of indices to contextualize findings across the indices.