The present study focused on event‐related potentials to tones in the presence and absence of sleep spindles. Six undergraduates were studied throughout an experimental night, following an adaptation session. The event‐related potentials to tone stimuli were averaged for each subject. Separate averages were determined for trials on which no sleep spindle occurred 2 s before or after a tone and trials in which spindle activity was present. Both voltage distribution maps and multivariate analysis of the waveforms produced significant differences between these conditions, which could be seen as a higher initial positive component and sustained positivity over the averaged epoch in the presence of spindles. Spectral analysis indicated that this result could not solely be ascribed to residual sigma activity in the spindle‐present average. The results may provide insights into the functional role of sleep spindles in humans in addition to that suggested by a neurophysiological model of inhibition.