Two experiments were performed in which the effects of selective spatial attention on the ERPs elicited by unilateral and bilateral stimulus arrays were compared. In Experiment 1, subjects received a series of grating patterns. In the unilateral condition these gratings were presented one at a time, randomly to the right or left of fixation. In the bilateral condition, gratings were presented in pairs, one to each side of fixation. In the unilateral condition standard ERP effects of visual spatial attention were observed. However, in the bilateral condition we failed to observe an attention related posterior contralateral positivity (overlapping the P1 and N1 components, latency interval about 100-250 ms), as reported in several previous studies. In Experiment 2, we investigated whether attention related ERP lateralizations are affected by the task requirement to search among multiple objects in the visual field. We employed a task paradigm identical to that used by Luck et al. (Luck, S.J., Heinze, H.J., Mangun, G.R., Hillyard, S.A., 1990. Visual event-related potentials index focused attention within bilateral stimulus arrays. II. Functional dissociation of P1 and N1 components. Electroencephalogr. Clin. Neurophysiol. 75, 528-542). Four letters were presented to a visual hemifield, simultaneously to both the attended and unattended hemifields in the bilateral conditions, and to one hemifield only in the unilateral conditions. In a focused attention condition, subjects searched for a target letter at a fixed position, whereas they searched for the target letter among all four letters in the divided attention condition (as in the experiment of Luck et al., 1990). In the bilateral focused attention condition, only the contralateral P1 was enhanced. In the bilateral divided attention condition a prolonged posterior positivity was observed over the hemisphere contralateral to the attended hemifield, comparable to the results of Luck et al. (1990). A comparison of the ERPs elicited in the focused and divided attention conditions revealed a prolonged 'search related negativity'. We discuss possible interactions between this negativity and attention related lateralizations. The display search negativity consisted of two phases, one phase comprised a midline occipital negativity, developing first over the ipsilateral scalp, while the second phase involved two symmetrical occipitotemporal negativities, strongly resembling the N1 in their topography. The display search effect could be modelled with a dipole in a medial occipital (possibly striate) region and two symmetrical dipoles in occipitotemporal brain areas. We hypothesize that this effect reflects a process of rechecking the decaying information of iconic memory in the occipitotemporal object recognition pathway.