In the attentional blink (AB), processing of a second target (T2) is impaired if it is presented shortly after the onset of a first target (T1), leading to a decrease in accurate report of T2 if T2 is masked. Some prominent theories of the AB suggest that an amodal bottleneck in working memory consolidation underlies the AB. We investigated this by factorially manipulating T1 and T2 modalities (visual or auditory) using equivalent stimuli and tasks in both modalities to minimize task switching. T2 was not masked. In all modality combinations, the electrophysiological P3 component to T2, obtained by subtracting T1 only trials from T1+T2 trials, was delayed and reduced in amplitude when T2 was presented soon after T1 relative to when T1 and T2 were presented farther apart. Results provide support for a common amodal bottleneck that underlies the AB effects observed in all visual/auditory modality combinations.