We have examined the behavioural consequences of a partial unilateral dopaminergic denervation of the rat striatum. This partial lesion was obtained by an intrastriatal 6-hydroxy-dopamine injection (6-OHDA, 20 or 10 microgram divided between two injection sites) and was compared with a unilateral complete lesion resulting from an injection of 6-OHDA (2 x 6 microgram) into the medial forebrain bundle. Quantification of striatal dopamine (DA) and its metabolites, and the immunohistochemical evaluation of the nigrostriatal DA system confirmed the complete and partial lesions. Animals with complete striatal denervation displayed both apomorphine- and amphetamine-induced rotations whereas the partial denervation elicited amphetamine-induced rotations only. However, the rates of amphetamine-induced rotation were not correlated with the size of the lesion. In contrast, the paw-reaching impairments were significantly correlated with the striatal dopaminergic depletion. When evaluated in the staircase test, animals with partial denervation were impaired exclusively for the paw contralateral to the side of the lesion. This motor deficit (50-75%) included all components of the skilled paw use (i.e. attempt, motor coordination and success) and was observed at least 12 weeks after the lesion. However, these animals were able to perform normal stepping adjustments with the impaired paw, indicating that the partial lesion induced a coordination deficit of the paw rather than a deficit of movement initiation. After a complete lesion, stepping adjustments of the contralateral paw were dramatically impaired (by 80%), an akinesia which almost certainly accounted for the great deficit in skilled paw use. The paw-reaching impairments resulting from the partial striatal denervation are proposed as a model of the early symptoms of Parkinson's disease and may be useful for the development of restorative therapies.