The availability of new data sources on human mobility is opening novel avenues for investigating the interplay of social networks, human mobility and dynamical processes such as epidemic spreading. We report on the design and deployment of a wireless sensor network for measuring human proximity in large-scale, real-world settings. We analyze networks of human contact measured in a variety of environments that include conferences, exhibitions, schools and hospitals. We characterize the empirical networks from both a static and a dynamic point of view, exposing important differences as well as striking similarities across the investigated environments, and we show the impact of network dynamics on spreading processes occurring over the contact network.