Globalisation and the knowledge-based society have driven universities to an intense competition for the best professors, researchers and students. Yet, rankings and reports measuring how universities perform are available in abundance. Among these assessment tools it is worth highlighting those based on Data Envelopment Analysis (DEA), which since the 1980s have risen considerably. However, there is a lack of consensus when selecting the indicators that best represent the inputs and outputs of such institutions. Aiming to clarify how these proxies are chosen, this paper does an exhaustive review of the indicators used in DEA empirical studies in the last decade, classifying them according to their nature and use.