Visual search for a target among distractors is often speeded when the target-defining feature is repeated from trial to trial, compared to when it changes. It has been proposed that whether this intertrial priming effect is perceptual in nature or not, and how strong it is, depends on the perceptual ambiguity of the display. Using the event-related potential (ERP) method, Experiment 1 assessed whether perceptual ambiguity indeed resulted in increased priming, by comparing target repetitions and changes in a condition with a target-resembling distractor to a condition without such a distractor. Priming effects on response times were indeed greater in the presence of a distractor. Moreover, EEG analyses revealed latency shifts and amplitude differences in the P1 and N2pc components for the same condition, consistent with a perceptual locus of intertrial priming. Experiment 2 confirmed this by varying the perceptual similarity of the distractor to the target. Priming benefits increased with increasing similarity. © 2009 Elsevier Ltd.