Several previous papers have suggested that the mismatch negativity (MMN), an event-related potential (ERP) component specific to stimulus change after repetition, is fully automatic, that is, not affected by attention. Woldorff, Hackley, and Hillyard (1991), however, report that the MMN elicited by an intensity decrement of a repeating stimulus is considerably attenuated when attention is strongly focused on other stimuli. The present commentary attributes most of this effect to attention effects on ERP components other than the MMN but also admits that some attentional reduction of the MMN component indeed occurred. This attenuation was interpreted by Woldorff et al. as indicating suppression of early sensory processing in unattended channels, but this is contraindicated by some very recent data from the present author's laboratory and elsewhere showing that the MMN to frequency change is not attenuated outside a strong attentional focus. Consequently, an alternative explanation is proposed for Woldorff et al.'s important finding.