When people attend to a target (called T1) in a sequence of rapidly presented visual items, their perception of a subsequent target is impaired if a) T1 is masked by an immediately succeeding item and b) the intertarget interval is less than half a second. This is the attentional blink. The blink is thought to reflect T1's attentional dwell time, the period over which it gains control over an end-of-trial response. I report several experiments in which I manipulated T1's ability to gain control of its response by varying the response compatibility of T1 and the item immediately following it. This research helps refine theoretical models of the attentional blink.