In Experiment 1 (change blindness), participants received either intermixed or blocked presentations of two visual stimuli that contained several common (X) and unique (A or B) features. On the critical trial after exposure, the stimulus AX was presented but included an unexpected visual event (a change in the size of a stimulus feature). We found that participants readily detected the change when it involved an A-unique feature that had been preexposed intermixed. However, if the change involved an A-unique feature preexposed in blocks, or an X-common feature (preexposed either intermixed or in blocks), the level of the detections considerably decreased. In Experiment 2, after either intermixed or blocked preexposure to AX and BX, all the participants were instructed to search for an unrelated visual target item. In the distractor condition, during the development of the search task, the A-unique feature emerged at an unpredictable location that was different to that occupied by this feature during the preexposure. In the control condition, the target was presented on its own during the entire task. The presence of the A-unique feature only produced a detrimental effect on search performance when the AX and BX had been preexposed intermixed. These results are discussed as indicative of the intermixed preexposure schedule enhancing the ability of the stimulus unique features to capture involuntary attention.