This paper discusses some problems associated with the concept of Pareto efficiency: first of all, the difficulties of defining what is a Pareto improvement and a Pareto optimum in social sciences (and economics in particular), and secondly, the relation between Pareto efficiency and equity of distributions of resources. In particular, we want to point out some ambiguities and contradictions of the concept of Pareto efficiency, which makes it hardly usable in real-world social contexts. When qualitative variables are involved, there are situations in which there is no straightforward way of deciding a preferable optimum within the Pareto set, other than taking each individual’s preference and choice into account when establishing what is and what is not an improvement (which is in many cases impossible) or having a ruling authority to establish that (which is undesirable). With this concept, we seem to be caught between the subjectivity of personal preference and the arbitrariness of dictatorship.