The learning sciences community's interest in learning analytics (LA) has been growing steadily over the past several years. Three recent symposia on the theme (at the American Educational Research Association 2011 and 2012 annual conferences, and the International Conference of the Learning Sciences 2012), organized by Paulo Blikstein, led to the meeting of learning scientists working in this area and ultimately generated the proposal for this special issue. In the two years that the authors have worked on putting together this special issue, the task of writing an introduction has become both much simpler and significantly more difficult. On the one hand, many of the trends that are driving the increasing attention to LA--big data, the Cloud--have become so prominent that the authors can count on readers to have some familiarity with them. Thus, the authors of this article do not need to start at the beginning of the discussion of LA for the "Journal of the Learning Sciences" ("JLS") audience. On the other hand, the scope of the field and the potential applications have grown tremendously in this short time. The result is that, if anything, the authors feel that they have fallen further behind. Although the educational data mining and LA communities have produced a steady stream of interesting results, work in education has far to go in order to reap the benefits for student learning that, say, businesses have reaped in advertising. The purpose of this special issue is to provide a focal point for researchers in learning sciences to understand and discuss the potential of LA. It is the hope of these authors that the special issue can introduce the community to the possibilities of LA and frame discussions of the future role of LA in the learning sciences.