© 2016 Nature America, Inc., part of Springer Nature. All rights reserved. NATURE MEDICINE VOLUME 22 | NUMBER 11 | NOVEMBER 2016 1 FOCUS ON PSYCHIATRIC DISORDERS PERSPECTIVE Neuropsychiatric disorders such as autism spectrum disorder (ASD), schizophrenia (SCZ) and bipolar disorder (BPD) are of great societal and medical importance, but the complexity of these diseases and the challenges of modeling the development and function of the human brain have made these disorders difficult to study experimentally. The recent development of 3D brain organoids derived from human pluripotent stem cells offers a promising approach for investigating the phenotypic underpinnings of these highly polygenic disorders and for understanding the contribution of individual risk variants and complex genetic background to human pathology. Here we discuss the advantages, limitations and future applications of human brain organoids as in vitro models of neuropsychiatric disease. Neuropsychiatric disorders, such as ASD, SCZ and BPD, have a devastating impact on patients’ lives and represent a heavy social and economic burden for societies worldwide 1,2 . The complex symptomatology of these diseases, which affect uniquely human cognitive and behavioral traits, combined with their polygenic etiology, the lack of objective biomarkers for diagnosis and the limitations of current research models, have made these disorders difficult to classify diagnostically and to study experimentally. Our current understanding of the structural, cellular and circuit-level abnormalities that affect the brains of patients with neuropsychiatric disorders remains very superficial, and the development of modern therapeutics has stalled 3 . Recent years, however, have witnessed unprecedented genomic studies of large cohorts of individuals with neuropsychiatric disor- ders, which has led to the current wealth of data on the genetics of these disorders 4,5 . With this has come the challenge of linking the rare, highly penetrant mutations and complex genetic variation identified in these studies to the neurobiological phenotypes reflective of disease states, and the need for faithful and tractable experimental models has taken center stage. It has become clear that whereas animal models have proven valuable for studying single-gene mutations, the majority of neuropsychiatric phenotypes involve heterogeneous combinations of many alleles of small effect, which are extremely difficult to recreate in animal models. In addition, conventional animal models, such as rodents, are limited by inherent species differences in the development, architecture and function of their brains. Primate models are also needed, including models of the human brain, but the use of endogenous human brain tissue is complicated by practical and ethical concerns of tissue availability, expansion and manipulation. Recent progress has enabled the development of cellular models of the developing human brain via the generation of 3D brain organoids derived from human pluripotent stem cells. Although reductionist in nature, brain organoids have great potential to comple- ment 2D cell culture models and animal studies to investigate aspects of human brain development and pathology. In this Perspective, we will focus on human pluripotent stem cell (hPSC)-derived in vitro models of neuropsychiatric disease and provide a discussion of the advantages, limitations and promise that 3D human brain organoids could have as new cellular platforms for investigating psychiatric disease origin and pathology, and, notably, for decoding the impact of genetic variants associated with mental illness on human brain development and function. Neuropsychiatric disease and the need for experimental models Neuropsychiatric disorders have been historically and are currently classified on the basis of clinical symptoms; however, there is considerable sharing of symptomatology and genetic risk across current disorder definitions, and each diagnostic category (ASD, BPD, SCZ, etc.) most likely represents a heterogeneous combina- tion of underlying conditions. Indeed, these disorders might be interrelated, as attested to by the high rate of comorbidity between them and the frequency with which individuals of a single pedigree and bearing the same risk alleles present with different clinically defined syndromes 6 . Neuropsychiatric disorders affect complex higher-order brain functions that involve multiple cell types across different brain regions, such as the prefrontal cortex, the thalamic reticular nucleus, the thalamus and the basal ganglia, among others. Disease pheno- types have been reported at scales ranging from gross architectural changes and abnormal activity patterns that might reflect altered cir- cuit function 7,8 to abnormal functioning of individual ion channels 9 . As the cellular, molecular and genetic phenotypes of these disorders are studied in ever-greater detail, it becomes increasingly apparent that individual clinically defined disorders show great heterogeneity across each of these modalities. The complexity of these disorders has impeded progress on elucidating their etiology and pathogenesis; consequently, our understanding of the structural, cellular and circuit-level abnormalities that affect the brains of patients with neuropsychiatric disorders remains superficial. Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, Massachusetts, USA. Correspondence should be addressed to P.A. (paola_arlotta@harvard.edu). Received 25 July; accepted 22 September; published online 26 October 2016; doi:10.1038/nm.4214 The promises and challenges of human brain organoids as models of neuropsychiatric disease Giorgia Quadrato, Juliana Brown & Paola Arlotta