This experiment assesses the variable amounts of attention that are required for a viewer to process two kinds of interruptions common to television: the shift from one message to a different, unexpected message and the reference to previously presented material that follows an interruption. Respondents viewed 24 television sequences composed of three segments: an initial segment drawn from one program, an interrupt segment drawn from a second program, and a reorient segment that continued the presentation from the first program. Initial and interrupt segment lengths were either 10 seconds or 30 seconds in length to produce a factorial combination of four message sequences. Attention to these interruptions was measured using reaction times to audio tones located 1 and 6 seconds after a shift. More attention was required to view the interrupt segments following 30-second initial sequences, especially at the 6-second tone location. For the reorient segments, sequences containing 30-second initial segments required more attention, as did 30-second interrupt segments. These results are interpreted in terms of limited capacity and attentional inertia models of attention.