This article examines conceptual, methodological, and analytic issues in research on relapse to alcohol and other drug use. The review notes the continued move in relapse research from a primary reliance on retrospective assessment of factors surrounding the onset of relapse episodes to an increased focus on the use of new technologies to obtain "near real time" data on proximal factors in relapses. Recent advances in neurobiology have yielded important gains in our understanding of vulnerability to relapse in alcohol and other drug abusers. New statistical techniques are also available to analyze data on relapse. From a theoretical standpoint, there has been an increasing appreciation for the complicated and dynamic interplay of distal and proximal factors in the relapse process. At this point, the strongest and most detailed data on factors in the onset and course of relapse have been generated by studies of smoking relapses that have made use of Ecological Momentary Assessment (EMA) technology. However, there is still limited "near real" time data on proximal factors in alcohol and other drug relapses, and no such data on factors that influence the course of these relapses, once they have begun. Nevertheless, important methodological advances have been and continue to be made in the study of relapse, and our knowledge about the nature and process of relapse has increased considerably over the past 10 years.