Observations of daily life are the bread and butter of ethnography but rarely feature as data in other kinds of work. Could non-ethnographic studies also benefit from such observations? If so, how? This article proposes ‘accidental ethnography’ as a method that field researchers can use to gain better understanding of the research context and their own social positioning within that context. Accidental ethnography involves paying systematic attention to the unplanned moments that take place outside an interview, survey, or other structured methods. In these moments the researcher might hear a surprising story or notice an everyday scene she had previously overlooked. The importance of these observations lies not in what they tell us about the particular, but rather what they suggest about the larger political and social world in which they (and the researcher) are embedded. The paper illustrates the argument by presenting five stories from the author’s experiences conducting research on local violence in Rwanda, Bosnia, the US, and elsewhere.