The fact that dopamine (DA) cells are lost specifically but other nerve cells are still intact makes Parkinson's disease (PD) uniquely amenable to cell replacement therapy. Early attempts have demonstrated that transplanted DA neurons derived from fetal tissue can indeed survive, reconnect with other nerve cells in the brain of PD patients and in many instances improve the clinical symptoms of the patient. However, the use of fetal tissue poses many technical, logistical, and ethical questions and it has become obvious that a renewable cell source is needed to allow many PD patients to benefit from this type of therapy. Human embryonic stem (ES) cells provide a potentially unlimited supply of specialized cells for regenerative medicine. However, a substantial challenge has been to control the types and numbers of therapeutically relevant cells generated from human ES cells in a predictable fashion.