When two masked targets (T1 and T2), both requiring attention, are presented within half a second of each other, report of the second target is poor, demonstrating an attentional blink (AB). Potter, Chun, Banks, and Muckenhoupt (1998) argued that all previous demonstrations of an AB occurring when one or more targets were presented outside the visual modality did not represent true AB but were, instead, artifactual, resulting from switching of task set. In the present experiments, T1 and T2 modalities were independent and varied randomly from trial to trial, allowing no useful preparatory task-set switching from T1 to T2. However, reliable ABs were observed when both targets were visual, when both targets were auditory, and cross-modally when T2s were visual. Furthermore, the ABs observed for cross-modality visual T2s showed the characteristic U-shaped pattern often found in AB experiments in which two visual targets are used--a pattern that should not be observed under task-set switching conditions. These results provide evidence that cross-modality AB can be found under conditions that do not allow useful preparatory task-set switching.