Animal models with partial lesions of the dopaminergic nigrostriatal pathway may be useful for developing neuroprotective and neurotrophic therapies for Parkinson's disease. To develop such a model, different doses of 6-hydroxydopamine (0.0, 0.625, 1.25, 2.5 and 5.0 mu g/mu l in 3.5 mu l of saline) were unilaterally injected into the striatum of rats. Animals that received 1.25 to 5.0 mu g/mu l 6-hydroxydopamine displayed dose-dependent amphetamine and apomorphine-induced circling. 6-Hydroxydopamine also caused dose-dependent reductions in [H-3]mazindol-labeled dopamine uptake sites in the lesioned striatum and ipsilateral substantia nigra pars compacta (up to 93% versus contralateral binding), with smaller losses in the nucleus accumbens, olfactory tubercle and ventral tegmental area. In the substantia nigra pars compacta and the ventral tegmental area, the number of Nissl-stained neurons decreases in parallel with the reduction in [H-3]mazindol binding. The reduction in [H-3]mazindol binding in the striatum and the nucleus accumbens, and the reduction in [H-3]mazindol binding and in the number of Nissl-stained neurons in the substantia nigra pars compacta and the ventral tegmental area is stable for up to 12 weeks after the lesion. Macroscopically, forebrain coronal sections showed normal morphology, except for rats receiving 5.0 mu g/mu l 6-hydroxydopamine in which striatal cross-sectional area was reduced, suggesting that this high dose non-specifically damages intrinsic striatal neurons. Nissl-stained sections revealed aa area of neuronal loss and intense gliosis centered around the needle track, which increased in size with thedoe of neurotoxin. Striatal [H-3]sulpiride binding was increased by 2.5 mu g/mu l and 5.0 mu g/mu l 6-hydroxydopamine, suggesting up-regulation of dopamine D-2, receptors. Striatal binding of [H-3]CGS 21680-labeled adenosine A(2a) receptors, but not of [H-3]SCH 23390-labeled dopamine D-1 receptors, was reduced at the highest dose, suggesting preservation of the striatal intrinsic neurons with the lower doses. This study indicates that intrastriatal injection of different doses of 6-hydroxydopamine can be used to cause increasing amounts of dopamine denervation, which could model Parkinson's disease of varying degrees of severity. Injecting 3.5 mu l of 2.5 mu g/mu l 6-hydroxydopamine appears to be particularly useful as a general model of early Parkinson's disease, since it induces a lesion characterized by robust drug-induced rotation, changes in binding consistent with similar to 70% dopamine denervation, similar to 19% dopamine D-2 receptor up-regulation, negligible intrinsic striatal damage and stability for at least 12 weeks. This study outlines a technique for inducing partial lesions of the nigrostriatal dopamine pathway in rats. These lesions model early Parkinson's disease, and can be useful for testing protective, preventive, restorative and symptomatic treatments.