This review by Karl J. Friston explores the evolving field of functional and effective connectivity in neuroimaging. Over the past two decades, neuroimaging has become a dominant technique in systems neuroscience, with a growing emphasis on understanding distributed processing and connectivity. The review is divided into three sections: a historical overview of functional integration, a discussion of pragmatic issues related to functional and effective connectivity, and an examination of recent advances in modeling experimental and endogenous network activity. The first section traces the historical development of functional integration, highlighting the shift from localizationism to connectionism. This shift has led to a greater focus on functional integration among segregated brain areas, which is crucial for understanding the brain's functional architecture. The second section delves into the practical aspects of functional and effective connectivity. It clarifies the distinction between these two concepts and explores the relationships among various analytic approaches. Functional connectivity is defined as statistical dependencies among remote neurophysiological events, while effective connectivity refers to the influence one neural system exerts over another. The operational distinction between these concepts is important for understanding the nature of inferences made about functional integration. The third section examines recent advances in modeling experimental and endogenous network activity, focusing on processing hierarchies and the distinction between forward and backward connections. It also discusses recent advances in network discovery and their application to hierarchical brain architectures. The review concludes by reflecting on the role of structural connectivity in providing constraints for effective connectivity analysis and the importance of generative models in understanding brain function. It emphasizes the need for a mechanistic understanding of disconnection syndromes and other disturbances of distributed processing, highlighting the potential of combining generative models with classification models to improve our understanding of brain disorders.This review by Karl J. Friston explores the evolving field of functional and effective connectivity in neuroimaging. Over the past two decades, neuroimaging has become a dominant technique in systems neuroscience, with a growing emphasis on understanding distributed processing and connectivity. The review is divided into three sections: a historical overview of functional integration, a discussion of pragmatic issues related to functional and effective connectivity, and an examination of recent advances in modeling experimental and endogenous network activity. The first section traces the historical development of functional integration, highlighting the shift from localizationism to connectionism. This shift has led to a greater focus on functional integration among segregated brain areas, which is crucial for understanding the brain's functional architecture. The second section delves into the practical aspects of functional and effective connectivity. It clarifies the distinction between these two concepts and explores the relationships among various analytic approaches. Functional connectivity is defined as statistical dependencies among remote neurophysiological events, while effective connectivity refers to the influence one neural system exerts over another. The operational distinction between these concepts is important for understanding the nature of inferences made about functional integration. The third section examines recent advances in modeling experimental and endogenous network activity, focusing on processing hierarchies and the distinction between forward and backward connections. It also discusses recent advances in network discovery and their application to hierarchical brain architectures. The review concludes by reflecting on the role of structural connectivity in providing constraints for effective connectivity analysis and the importance of generative models in understanding brain function. It emphasizes the need for a mechanistic understanding of disconnection syndromes and other disturbances of distributed processing, highlighting the potential of combining generative models with classification models to improve our understanding of brain disorders.