Therapeutic development and research in the neurodegenerative disease field encounters many challenges such as availability of reproducible and scalable cellular model systems that are biologically, physiologically, and pharmacologically relevant. These cellular models must be informative of cellular mechanisms of diseases and predictive for therapeutics efficacy and toxicity testing during drug discovery and development. Neural spheroids fill the gap of cellular models of the brain that are functional, versatile in neural cell type composition, robust, and scalable for high-throughput screening (HTS). We have previously developed a protocol to aggregate pre-determined ratios of differentiated human-induced pluripotent stem cell (hiPSC)-derived neurons and astrocytes in a scaffold-free environment to form 3D brain-region specific spheroids. By mixing different neuronal types, neural spheroids can be used to simulate the neuronal-type heterogeneity of distinct brain regions in vivo, including the prefrontal cortex (PFC) and ventral tegmental area (VTA). Here, we present a detailed description of a method for generating functional brain region-specific spheroids with HTS-compatible assay readout that monitors changes in neural network activity by measuring calcium oscillations. The versatility of the platform is such that these neural spheroids cellular assays are applicable for a wide range of disease modeling, compound validation, and screening and are limited only by the availability of input cells, including neural subtype, disease cells, and immune cells such as microglia.