The assumption that children born preterm have difficulties in maintaining active attention was tested in passive and active tasks. Twenty 5-year-old children born preterm at 26 to 32 weeks gestational age were compared with 20 children born at term, matched for age and IQ, using an auditory paradigm. In the passive task participants had to watch a videotape of a cartoon and ignore auditory stimuli. In the active task they had to detect a rare tone (the ‘target’ tone; 10% of the tones presented) among frequent tones (the ‘standard’ tone; 90%). Accuracy and reaction time were analysed, and event-related potentials (ERPs) were recorded at the scalp sites Fz, Cz, T3, T4, Pz, Oz, and two electrodes for the left mastoid (passive task); and Fz, F7, F8, Cz, T3, T4, Pz, and Oz (for the active task). Behavioural and electrophysiological data were analysed with repeated-measure ANOVAs. The results showed a significant group effect only on the active task. The preterm group scored fewer correct hits (correct detection of target tone) and were less efficient in their attentional strategy as assessed by ERP components.