The following question arises in flame-cutting and similar applications. "Given a graph drawn in the plane, is there an Eulerian circuit in which successive edges always belong to a common face?" We prove that this question and related ones are NP-complete