Multiple system atrophy (MSA) is a neurodegenerative disorder that occurs sporadically and causes parkinsonism, cerebellar, autonomic, urinary, and pyramidal dysfunction in many combinations. Progressive L-dopa-unresponsive parkinsonism due to underlying striatonigral degeneration dominates the clinical syndrome in the majority of cases (MSA-P subtype). MSA-P is characterized pathologically by degenerative changes in somatotopically related areas of the substantia nigra pars compacta and of the putamen. Furthermore, oligodendroglial cytoplasmic inclusions (GCIs) are observed throughout the cortico-striato-pallidocortical loops and may contribute to the basal ganglia dysfunction. Neurotransplantation strategies are of potential interest in this disease, which causes marked and early disability and dramatically reduces life expectancy. A number of experimental MSA-P models have been employed to evaluate neurotransplantation approaches. Sequential nigral and striatal lesions using 6-hydroxydopamine and quinolinic acid (double toxin-double lesion approach) indicate that apomorphine-induced contralateral rotation is abolished by a secondary striatal lesion. Intrastriatal injection of mitochondrial respiratory chain toxins produces secondary excitotoxic striatal lesions combined with retrograde nigral degeneration and therefore provides an alternative single toxin-double lesion approach. Neurotransplantation in MSA-P animal models has been used to improve functional deficits by replacing lost nigral and/or striatal circuitry (neuroregenerative approach). The available data indicate that embryonic mesencephalic grafts alone or combined with striatal grafts partially reverse drug-induced rotation asymmetries without improving deficits of complex motor function. The potential neuroprotective efficacy of embryonic striatal grafts against striatal excitotoxicity is presently under investigation in the double toxin-double lesion MSA-P rat model. Anecdotal clinical evidence in one MSA-P patient misdiagnosed as Parkinson's disease indicates that embryonic mesencephalic grafts produce incomplete clinical benefit. Striatal co-grafts may increase functional improvement. Further experimental studies are required prior to the clinical application of embryonic neurotransplantation in MSA-P. Future research strategies should explore the effect of neurotransplantation in partial MSA-P rat models with less severe nigral and striatal degeneration, the feasibility of a primate model closely mimicking the human disease, and the replication of oligodendroglial dysfunction.