Neuropsychiatric disorders which target defined neural circuits in the brain have been difficult to treat, due to the complexity of the human brain and inaccessibility to tissue for study. As a result, there has been a dearth of successful pharmaceutical therapies available to treat these disorders, which is largely due to inadequate predictive models of efficacy. Human iPSC-derived neurons hold promise as one such system, but to date have been primarily used in single neuronal population studies focused toward unraveling complex neurological disease mechanisms and genetic studies. While this is an important step toward the goal of better therapeutic development, these cell cultures have been removed from their neural circuit contexts, which may also remove important disease endophenotypes which can only be seen in circuit models. We have therefore developed microfluidic-based circuit models using human neurons which can capture certain aspects of these neural circuits. These microdevices are designed for compartmentalization of distinct neuronal subtypes, resembling the connections between varying brain nuclei. Here, we demonstrate that human induced neurons establish functional synaptic contacts and exhibit circuit network activity within and between device compartments. These circuits can be excited by small molecules or optogenetics and altered using receptor antagonists, and their functional activity measured by either patch clamp or calcium imaging. Furthermore, we have developed additional computational tools which aid in the processing of large amounts data, thereby accelerating the interpretation of data and facilitating high-throughput screening studies.