The neural mechanisms that underlie both the anti-parkinsonian effects of dopamine agonists and dopamine agonist-induced dyskinesia were studied in parkinsonian primates, using the regional brain uptake of [3H]2-deoxyglucose (2-DG). Parkinsonian symptoms were induced in monkeys by the administration of the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). Some of the animals received chronic dopamine replacement therapy for at least 3 mths, which resulted in the appearance of peak-dose dyskinesia. The remaining animals spent an equivalent period of time relatively unexposed to dopaminergic agents, receiving only therapeutic doses, and at no time showed any signs of dopamine agonist-induced dyskinesia. The 2-DG metabolic mapping technique was applied to all of these animals shortly following the administration of a dose of dopamine agonist which was sufficient to alleviate their parkinsonian symptoms and to induce dyskinesia in those prone to this complication. The 2-DG uptake technique permits the autoradiographic measurement of local cerebral glucose uptake which was used as an index of regional synaptic activity. The resultant autoradiographs were compared with those from a previous study which examined 2-DG uptake in parkinsonian and normal brains from animals which had not received dopamine agonists prior to the terminal 2-DG uptake procedure. The pattern of 2-DG uptake in the animals which received a dopamine agonist prior to the terminal procedure was strikingly different to both of the other groups. The most affected structure was the subthalamic nucleus which showed a dramatic increase in 2-DG uptake in animals exposed to dopamine agonist immediately prior to the terminal procedure, especially in the ventral tip of the nucleus. The medial pallidal segment also showed relatively greater levels of 2-DG uptake in the dopamine agonist group compared with the untreated parkinsonian state whereas, in contrast, the lateral pallidal segment showed decreased levels of 2-DG uptake. The parkinsonian animals which had been sacrificed after receiving a dopamine agonist were split into three groups on the basis of their response to the agonist. The first group had their parkinsonism reversed and appeared clinically normal. The remaining animals had their parkinsonism reversed by the dopamine agonist but showed dyskinesia at peak dose. These animals were allotted to two further groups depending on whether their dyskinesia was of a choreic or dystonic nature.(ABSTRACT TRUNCATED AT 400 WORDS)