The non-human primate models of Parkinson's disease which have been developed using the neurotoxin MPTP (1-methyl-4-phenyl-1,2,3,6 tetrahydropyridine) have proven to be either unstable or variable, or to display only a limited subset of parkinsonian features. The present study examined a new two-stage lesion approach in which MPTP was administered via the carotid arteries. The first infusion through one artery produced a hemiparkinsonian state and was followed several months later by a second MPTP infusion into the contralateral carotid artery to induce bilateral parkinsonism. Animals receiving lesions were evaluated using a battery of tests which included a monkey parkinsonism rating scale, a movement time-task and continuous monitoring of home cage activity. All animals monitored showed significant decreases in activity levels of up to 95% following the second lesion. These decreased activity levels remained stable throughout the observation period of up to 12 months postlesion. In addition to the decreased home cage activity, bilaterally lesioned animals displayed bilateral parkinsonian features including akinesia, bradykinesia, rigidity, tremor and balance and gait disturbances which were stable, following an acute period of up to 45 days, for the remainder of the study. Administration of levodopa increased activity levels and reduced motor dysfunctions. Thus, a two-stage bilateral lesion approach, utilizing the neurotoxin MPTP, appears to provide a less variable and relatively stable model of bilateral Parkinson's disease in nonhuman primates. Treated animals display the cardinal features of parkinsonism and respond appropriately to the standard antiparkinsonian drug, levodopa.