This paper deals with data dissemination in resource-constrained opportunistic networks, i.e., multi-hop ad hoc networks in which simultaneous paths between endpoints are not available, in general, for end-to-end communication. One of the main challenges is to make content available in those regions of the network where interested users are present, without overusing available resources (e.g., by avoiding flooding). These regions should be identified dynamically, only by exploiting local information exchanged by nodes upon encountering other peers. To this end, exploiting information about social users’ behaviour turns out to be very efficient. In this paper we propose and evaluate ContentPlace, a system that exploits dynamically learnt information about users’ social relationships to decide where to place data objects in order to optimise content availability. We define a number of social-oriented policies in the general Content-Place framework, and compare them also with other reference policies proposed in the literature. Copyright 2008 ACM.