Although measuring income inequality seems straightforward and uncontroversial, methodological issues greatly affect findings. This chapter shows that changes in real median income between 1979 and 2014 across six studies varied from negative 8 percent to positive 51 percent. Furthermore, the share of growth going to the top ten percent during these years in four studies ranged from 31 percent to 100 percent. The first choice that researchers make is choosing a dataset or linked datasets. This choice affects the income sharing unit, be it households, families, individuals, or tax units. The next choice is the definition of income, with the starting point being cash market income only—earnings, dividends, rents, interest payments, or business profits. This total can be expanded by including government cash benefits, employer benefits, the rental value of home ownership, and government and financial services that people don’t pay for. Even after the income concept is chosen, income can be presented as adjusted for family size and either before or after taxes. Finally, adjusting for inflation to change nominal incomes into inflation-adjusted incomes can be performed in a variety of ways.