The aim of this study was to address the potential for the auditory mismatch negativity (aMMN) to be used in applied event-related potential (ERP) studies by determining whether the aMMN would be an attention-dependent ERP component and could be differently modulated across visual tasks or virtual reality (VR) stimuli with different visual properties and visual complexity levels. A total of 80 participants, aged 19–36 years, were assigned to either a reading-task (21 men and 19 women) or a VR-task (22 men and 18 women) group. Two visual-task groups of healthy young adults were matched in age, sex, and handedness. All participants were instructed to focus only on the given visual tasks and ignore auditory change detection. While participants in the reading-task group read text slides, those in the VR-task group viewed three 360° VR videos in a random order and rated how visually complex the given virtual environment was immediately after each VR video ended. Inconsistent with the finding of a partial significant difference in perceived visual complexity in terms of brightness of virtual environments, both visual properties of distance and brightness showed no significant differences in the modulation of aMMN amplitudes. A further analysis was carried out to compare elicited aMMN amplitudes of a typical MMN task and an applied VR task. No significant difference in the aMMN amplitudes was found across the two groups who completed visual tasks with different visual-task demands. In conclusion, the aMMN is a reliable ERP marker of preattentive cognitive processing for auditory deviance detection.