Human learning involves flexibility in adapting brain function and precision in selecting new neurophysiological activities to drive desired behavior. These attributes operate across multiple temporal scales as skill performance transitions from slow and challenging to fast and automatic. Modular structure naturally supports selective adaptability, playing a critical role in evolution, development, and optimal network function. Using functional connectivity measurements from initial training through mastery of a simple motor skill, the study investigates the role of modularity in human learning by identifying dynamic changes in modular organization across multiple temporal scales. Results show that flexibility, measured by node allegiance to modules, in one session predicts learning in a future session. A general statistical framework for identifying modular architectures in evolving systems is developed, applicable to disciplines where network adaptability is crucial. The brain is a complex system that dynamically adapts to a changing environment over multiple temporal scales. Rapid adaptation and continuous evolution of interactions form the neurophysiological basis for behavioral adaptation or learning. Stable neurophysiological signatures of learning are best demonstrated in animal systems at the level of individual synapses. At larger spatial scales, specific regional changes in brain activity and effective connectivity accompany many forms of learning in humans, including motor skill acquisition. Learning-associated adaptability is thought to stem from the principle of cortical modularity. Modular structures are aggregates of small subsystems that can perform specific functions without perturbing the remainder of the system. Such structure provides compartmentalization and redundancy, reducing interdependence of components, enhancing robustness, and facilitating behavioral adaptation. Modular organization also confers evolvability on a system by reducing constraints on change. A putative relationship between modularity and adaptability in the context of human neuroscience has recently been posited. However, the existence of modularity in large-scale cortical connectivity during learning has not been tested directly. Based on theoretical and empirical grounds, the study hypothesized that modularity characterizes the fundamental organization of human brain functional connectivity during learning. Based on studies relating the neural basis of modularity to skilled movement development, it was expected that functional brain networks derived from a simple motor skill would display modular structure across the variety of temporal scales associated with learning. It was also hypothesized that modular structure would change dynamically during learning, and that characteristics of such dynamics would be associated with learning success. These predictions were tested using functional magnetic resonance imaging (fMRI), an indirect measure of local neuronal activity, in healthy adult subjects during the acquisition of a simple motor learning skill composed of visually cued finger sequences. Low-frequency functional networks were derived from fMRI data by computing temporal correlations between activity in each pair of brain regions to construct weighted graphs or whole-brain functional networks. This network framework enabled estimation of a mathematical representation of modular or community organization, known as 'network modularity', for each individual over a range of temporal scales. The evolution of network connectivity over time was evaluated using a novel mathematical framework, and its relationship with learning was tested using the Pearson correlation coefficient. Results showed that network organization was significantly modular atHuman learning involves flexibility in adapting brain function and precision in selecting new neurophysiological activities to drive desired behavior. These attributes operate across multiple temporal scales as skill performance transitions from slow and challenging to fast and automatic. Modular structure naturally supports selective adaptability, playing a critical role in evolution, development, and optimal network function. Using functional connectivity measurements from initial training through mastery of a simple motor skill, the study investigates the role of modularity in human learning by identifying dynamic changes in modular organization across multiple temporal scales. Results show that flexibility, measured by node allegiance to modules, in one session predicts learning in a future session. A general statistical framework for identifying modular architectures in evolving systems is developed, applicable to disciplines where network adaptability is crucial. The brain is a complex system that dynamically adapts to a changing environment over multiple temporal scales. Rapid adaptation and continuous evolution of interactions form the neurophysiological basis for behavioral adaptation or learning. Stable neurophysiological signatures of learning are best demonstrated in animal systems at the level of individual synapses. At larger spatial scales, specific regional changes in brain activity and effective connectivity accompany many forms of learning in humans, including motor skill acquisition. Learning-associated adaptability is thought to stem from the principle of cortical modularity. Modular structures are aggregates of small subsystems that can perform specific functions without perturbing the remainder of the system. Such structure provides compartmentalization and redundancy, reducing interdependence of components, enhancing robustness, and facilitating behavioral adaptation. Modular organization also confers evolvability on a system by reducing constraints on change. A putative relationship between modularity and adaptability in the context of human neuroscience has recently been posited. However, the existence of modularity in large-scale cortical connectivity during learning has not been tested directly. Based on theoretical and empirical grounds, the study hypothesized that modularity characterizes the fundamental organization of human brain functional connectivity during learning. Based on studies relating the neural basis of modularity to skilled movement development, it was expected that functional brain networks derived from a simple motor skill would display modular structure across the variety of temporal scales associated with learning. It was also hypothesized that modular structure would change dynamically during learning, and that characteristics of such dynamics would be associated with learning success. These predictions were tested using functional magnetic resonance imaging (fMRI), an indirect measure of local neuronal activity, in healthy adult subjects during the acquisition of a simple motor learning skill composed of visually cued finger sequences. Low-frequency functional networks were derived from fMRI data by computing temporal correlations between activity in each pair of brain regions to construct weighted graphs or whole-brain functional networks. This network framework enabled estimation of a mathematical representation of modular or community organization, known as 'network modularity', for each individual over a range of temporal scales. The evolution of network connectivity over time was evaluated using a novel mathematical framework, and its relationship with learning was tested using the Pearson correlation coefficient. Results showed that network organization was significantly modular at