While attending the last three AARE conferences I have noticed the theme of inequality running through many of the papers. It occurred to me that a short methodological paper on the measurement of inequality in education might be useful for educational researchers. Inequality is said to take place when a resource is distributed among individuals or among groups in such a way that we feel that some individuals or groups have got more than their "fair" share and others have got less. Resource is used here to mean something with a positive value, for example annual income or higher education places. Groups here can be defined in a variety of ways, for example by socioeconomic status or by geographical location. The term "fair" is more difficult to grasp; it brings in notions of social justice or equity. Strictly speaking we should separate the terms equality and equity, reserving equality for the more mathematical concept and equity for the concept of fairness or justice that underlies the concept of equality. In practice, and in keeping with the literature in this area, I shall sometimes use the two terms interchangeably. I believe this is an indication of how closely bound together the two concepts are. Most of us can recognise inequality when we see it and can readily make pronouncements like "such inequality is intolerable" or "this is grossly unfair". Why therefore do we need to measure inequality? There are at least two reasons. The first reason is that we are also very readily inclined to make pronouncements along the lines of "this inequality is much worse than it was ten years ago" or "inequality in Queensland is much greater than inequality in Utopia". Here we are making comparisons. It is therefore desirable that we get a parsimonious description, preferably in the form of a single numerical index, of inequality in one situation and a similar description of inequality in the other situation. Then when compare two unequal situations, as we did in the foregoing pronouncements, we can do so with some confidence and perhaps without ambiguity. The second reason for measuring inequality is that it forces us to have very clear concepts of what we mean by inequality and at the same time it forces us to examine our underlying concepts of equity. While reducing an inequality situation to a single numerical index may seem a very mechanistic approach, I would argue that not to do so when it is appropriate to do so is to invite very sloppy thinking on just what is meant by inequality. I would also argue that to accept inequality measures with blind faith without realising their limitations is equally indicative of sloppy thinking.