While many anthropologists have previously refl ected on longitudinal ethnography— for example distinguishing between diff erent categories of longitudinal research, including the ethnographic revisit, either by the same or another researcher, diachronic research projects, involving continuous and sustained engagement over time, or so-called large-scale or multigenerational projects, among others—there has been little refl ection on the way particular topics of research might impact on the longitudinal research process. In particular, we argue here that the stakes of longitudinal ethnographic research come to the fore particularly starkly in relation to studies of violence. More specifi cally, longitudinality potentially both enhances certain risks inherent to carrying out research on violence, while also off ering unique opportunities for better understanding the phenomenon more refl exively.