Seam effects are the tendency for estimates of change measured across the seam between two successive waves to far exceed change estimates measured within a single wave. They are present in longitudinal surveys for many behavioral variables, independently of the data collection technique or the length of the recall period. The seam problem is a quite recent issue; in survey research it has been studied mainly over the last 25 years. The first papers on this topic assessed the magnitude of seam bias and attempted to precisely measure its impact using validation data. The next efforts were directed toward understanding its possible causes. In the past few years, methods attempting to reduce seam bias have been tested and some have proven successful. Thus far, however, no strategy has been shown to completely eliminate the problem. This article summarizes the current literature on seam bias and concludes with suggestions for a future research agenda.