Abstract CP-networks have been proposed as a simple and intuitive graphical tool for representing conditional ceteris paribus preference statements over the values of a set of variables. While the problem of reasoning with CP-networks has been receiving some attention, there are very few works that address the problem of learning CP-networks. In this work we investigate the task of learning CPnetworks, given access to a set of pairwise comparisons. We fifirst prove that the learning problem is intractable, even under several simplifying assumptions. We then present an algorithm that, under certain assumptions about the observed pairwise comparisons, identififies a CP-network that entails these comparisons. We fifinally show that the proposed algorithm is a PAC-learner, and, thus, that the CP-networks it induces accurately predict the user’s preferences on previously unseen situations