Rats were trained in a specially designed apparatus to detect brief, unpredictable visual stimuli presented to either side of the head. In one condition, rats reported the detection of the visual stimulus by removing their heads from a central location and responding in one of two adjacent side-holes where the visual stimulus had occurred. In the other condition, rats were trained to respond in the hole on the opposite side to where the visual stimulus had occurred. Following training all rats received striatal infusions of 6-hydroxydopamine leading to profound striatal dopamine depletion. One group received unilateral intra-nucleus accumbens 6-hydroxydopamine infusions. Two groups received 6-hydroxydopamine unilaterally into the caudate nucleus. Two other groups received two infusions of 6-hydroxydopamine, intra-accumbens and intra-caudate, on either the same side or on opposite sides. The results showed that all groups, except that receiving only intra-accumbens 6-hydroxydopamine, exhibited a significant bias in responding towards the side of the lesion which correlated with dopamine depletion in the head of the caudate nucleus, regardless of the type of discrimination or pre-operative strategy. These biased groups were also slower to initiate, but not to complete, contralateral responses. Detailed analysis of the behavioural evidence suggested that unilateral striatal dopamine depletion does not produce sensory loss or sensory inattention but rather an output type neglect, perhaps related to hemiakinesia or "intentional neglect". The bias recovered to preoperative levels by 7 weeks after surgery, but could be reinstated by pretreatment with alpha-methyl-para-tyrosine, thus suggesting the involvement of presynaptic mechanisms in recovery. Although unilateral dopamine depletion from the nucleus accumbens alone had no effects on visual neglect, it produced a lateralized bias in premature responding away from the side of the lesion, contrasting with the bias towards the side of the lesion seen in all of the other groups. These effects appeared to summate in this and in another condition in which a more eccentric response was required for stimulus detection, to suggest an interaction in function between the dorsal and ventral striatum.