The current study investigated whether task-relevant information affects performance differently from how information that is not relevant for the task does when presented peripherally and centrally. In three experiments a target appeared inside the focus of attention, whereas a to-be-ignored distractor appeared either in the periphery (Experiments 1 and 2) or at the centre (Experiment 3) of attention. In each trial the distractor carried both task-relevant and irrelevant information. The results confirmed the "task relevance" hypothesis: Task-irrelevant information affected performance only when it appeared at the centre of attention, whereas task-relevant information affected performance when it appeared inside as well as outside the main focus of attention. The current results do not support suggestions that spatial stimuli (e.g., arrows) draw attention automatically regardless of task relevance.