This thesis presents income, earnings and consumption inequality data among Brazilian families. Those data were generated from two different sources: PNDA’s (National Household Survey) micro data for Brazil from 1976 to 1997 and POF’s (Household Expenditures Survey) micro data for the Brazilian metropolitan areas in 1987/88 and 1995/96. Theil’s indexes for family income and earnings among families headed by individuals with same age and educational level were calculated from the PNAD’s data for each year. The same procedure was used for POF’s data, but measures of consumption inequality were also done. From PNAD’s data, one can identify the part of inequality that is explained by age, cohorts, and time effects. The same sort of hypothesis tha Deaton and Paxson (1994) used to identify those three different effects was applied in this current work. In general, the consumption inequality tends to be a more accurate estimator of the permanent disparity of resources and of well being among families than the income inequality is. Therefore, calculations of consumption inequality were held using POF’s data. According to the permanent income hypothesis (PIH), one should expect that the consumption inequality among families of the same cohort grows as time evolves. In the case that there are some constraints to the PIH, as prudent consumers or credit constraints, the consumption inequality stars depending on the temporal evolution of income and earnings distribution, which allows that consumption inequality does not grow with time.