The mismatch negativity, an ERP that reflects the detection of change in the auditory environment, is considered to be a relatively automatic process. Its automaticity has by in large been studied using the oddball paradigm, in which a physical feature of a frequently presented standard stimulus is changed. In the present study, the automaticity of the MMN is tested using a MMN elicited by a violation of a more abstract auditory pattern. Fourteen subjects were presented with an alternating pattern of two tones (ABABAB) that was occasionally broken by deviant repetitions (e.g., ABABABBBAB). The alternating tones were separated by 1 or 6 semitones in different conditions. The subjects were engaged in a continuous multiple object tracking (MOT) task and thus ignored the auditory stimuli. Difficulty of the MOT task was manipulated by increasing the number of objects to be tracked. Subjects were also asked to read a text and ignore the auditory stimuli in another condition. A much larger MMN was elicited by pattern violations in the 6 than in the 1 semitone condition. The difficult visual task should have presumably required greater attentional focus than the easy task, and performance did deteriorate during the difficult MOT. The MMN, however, was not affected by the demands of the MOT task. This finding suggests that the MMN elicited by the violation of a pattern is not affected by the presumed attentional demands of a difficult continuous task such as multiple object tracking.