Embryonic stem cells (ESCs) provide hope as a potential regenerative therapy for neurological conditions such as Parkinson's disease and spinal cord injury. Currently, ESC-based nervous system repair faces several problems. One major hurdle is related to problems in generating large and defined populations of the desired types of neurons from human ESCs (hESCs). Moreover, survival of grafted hESC-derived cells has varied and functional recovery in recipient animals has often been disappointing. Importantly, in clinical trials, adverse effects after surgery, including tumors or vigorous immune reactions, must be avoided. Here we highlight attempts to overcome these hurdles with hESCs intended for central nervous system repair. We focus on hESC-derived dopamine-producing neurons that can be grafted in Parkinson's disease and identify critical experiments that need to be conducted before clinical trials can occur.