Clinical studies involving intrastriatal transplantation of embryonic mesencephalic tissue in patients with Parkinson disease (PD) have provided proof-of-principle for the cell replacement strategy in this disorder. The grafted dopaminergic neurons can reinnervate the denervated striatum, restore regulated dopamine release and movement-related frontal cortical activation, and produce significant symptomatic relief. In the most successful cases, patients have been able to withdraw from levodopa treatment after undergoing transplantation and resume an independent life. There are, however, several problems linked to the use of primary embryonic tissue: 1) lack of sufficient amounts of tissue for transplantation in a large number of patients; 2) variability of functional outcome (major improvement in some and modest if any clinical benefit in others); and 3) occurrence of troublesome dyskinesias in a significant proportion of patients after transplantation. Thus, neural transplantation is still at an experimental stage in the treatment of PD. For the development of a clinically useful cell therapy we need to define better criteria for patient selection and how graft placement should be optimized in each individual. Most importantly, we need to generate large numbers of viable dopamine neurons in preparations that are standardized and quality controlled. Stem cells could be useful as an unlimited source of dopamine neurons. Thus far, neurons with at least some dopaminergic characteristics have been generated from stem cells. In most cases, however, their survival after grafting in animals has been poor, and it is also unclear if they function as normal dopamine neurons. Several scientific issues need to be addressed before stem cell-based therapies can be tested in PD patients.