Conducting methodologically rigorous, community epidemiologic studies of the outcomes of large-scale potentially traumatic events (PTEs) presents extraordinary challenges. The primary scientific and logistical challenges of conducting such studies, however, result primarily from the fundamental unpredictability of the exposures of interest (i.e., the fact that disasters typically occur with little or no warning). Most of the important design problems arise from two unavoidable factors: the necessarily observational nature of the studies, and the need for the studies to be implemented quickly. In this chapter, we describe how using the Internet can enhance the design of these studies, focusing on three critical design features: probability sampling; psychometrically sound assessment of outcomes and other postexposure factors; and the use of prospective, longitudinal designs. We also describe recently developed methods for conducting Internet-based community epidemiologic studies, including prerecruited participant panels and multiple mode designs. We conclude that the Internet can be a valuable tool in conducting these studies, but that it will not be feasible as an assessment mode for some kinds of PTEs. Rather, Internet-based assessment is best understood as one of many tools that investigators will need in their tool kit for conducting scientifically sound, cost-efficient community epidemiologic studies of outcomes associated with large-scale PTEs.