This investigation examines 'teasing' of non-combatant children by US military service personnel in occupied Iraq and Afghanistan. The majority of existent investigations of teasing and related practices place significant conceptual importance on the intentions of the teaser - such that a target can understand that the tease is not true. However, in data examined here it appears that targets (children) do not understand the language in which the teasing is undertaken. Drawing from publicly available video footage posted on the video sharing website Liveleak, we provide an ethnomethodological (e.g., Garfinkel 1967) and conversation analytic informed (e.g., Sacks, Schegloff and Jefferson 1974; Schegloff 2007) examination of how persons initiating teasing (soldiers) strategically exploit asymmetries in the sequential and preferential organization of interactions when tease recipients (children) do not have sufficient English skills to redress (or understand) the negative assessments being made of them. Three types of candidate teasing practices are identified: soldier initiated negative other-assessments; target parroting negative other-assessments; and offer-withdrawal games. Analysis examines how such interactions effectively fail as teases and explores how children can resist soldiers' pursuit of degrading responses.