<p>The present experiments measure a form of preparatory control over selective attention in a two-target task. We used a skeletal two-target method in which the first of two target words (T1) is presented with or without an accompanying distractor word. When T1 is displayed with a distractor, a stark identification deficit (an attentional blink) is produced for the second target word (T2). In our adaptation of this method, we inserted a stimulus immediately prior to each two-target trial to measure sequence effects on selective attention to T1, and the resulting two-target cost for T2. Generally speaking, neither T1 nor T2 identification were influenced by the stimulus/response features of the event that preceded T1. However, both T1 and T2 identification were influenced by the task instructions for the event that preceded T1. These results point to a strong contribution of preparatory state on selective attention in a two-target task.</p>