The lack of relevant in vitro neural models is an important obstacle on medical progress for neuropathologies. Establishment of relevant cellular models is crucial both to better understand the pathological mechanisms of these diseases and identify new therapeutic targets and strategies. To be pertinent, an in vitro model must reproduce the pathological features of a human disease. However, in the context of neurodegenerative disease, a relevant in vitro model should provide neural cell replacement as a valuable therapeutic opportunity. Such a model would not only allow screening of therapeutic molecules but also can be used to optimize neural protocol differentiation [for example, in the context of transplantation in Parkinson's disease (PD)]. This study describes two in vitro protocols of 1) human glioblastoma development within a human neural organoids (NO) and 2) neuron dopaminergic (DA) differentiation generating a three-dimensional (3D) organoid. For this purpose, a well-standardized protocol was established that allows the production of size-calibrated neurospheres derived from human embryonic stem cell (hESC) differentiation. The first model can be used to reveal molecular and cellular events occurring during in glioblastoma development within the neural organoid, while the DA organoid not only represents a suitable source of DA neurons for cell therapy in Parkinson's disease but also can be used for drug testing.