Our era has started to be known as the Information Age. This name reflects the importance that fast communication means and information retrieval tools as Internet and the WWW are gaining in our everyday life. Since the opening of Internet to the general public, an important question is whether it is possible to predict the traffic that users generate in Web sites. The answer to this question, and most importantly a reliable method to do such prediction, could have immediate practical consequences. Examples are PageRank and the search engine that it inspired (Google), but also guiding automatic search processes (crawlers) or predicting advertising revenues for the sites. In this talk, I will describe our efforts to bridge the gap between real data and models in this area. We have performed several data collection campaigns with the aim of tracking navigation patters of users. Each individual user has his/her own particular characteristics, but we have found some common statistical features underlying their behavior in the Web. This allows us to propose realistic models able to reproduce individual Web surfing and by the aggregation of the different users to study site traffic.