Brain-machine interfaces, including electrical, magnetic and light stimulations are great challenges with appealing perspectives for engineering, ethics, neuroscience, neurology, and neuropsychiatry. Functional neuroimaging studies have highlighted the neuroanatomical perspective in brain and mental disorders, more specifically with therapeutical focal stimulation of neural networks. Many different brain stimulation techniques have successfully been developed to relieve pharmaco-resistant patients suffering of severe neurological (e.g, Parkinson's disease and neurogenic pain) and psychiatric disorders (e.g, obsessive-compulsive disorders, Gilles de la Tourette syndrome and schizophrenia). However, there are many knowns and unknows when it comes to understand the mechanisms underlying the therapeutic actions and side effects of these stimulation procedures. For instance, high-frequency electrical stimulation of the subthalamic nucleus impacts all nearby and remote structures having anatomical and/or functional links with it. Subthalamic stimulation alleviates the motor disorders of patients with advanced Parkinson's disease and can also, in some of them, modulate some facets of their behavior and personality. In parallel, in order to understand the brain functioning under physiological and pathological conditions, various cell-to-network electrophysiological techniques have been designed for the anatomofunctional exploration of brain circuits. Nerve cells are individually endowed with extraordinary underestimated performances, leading them with the power to control brain states and behavior. Indeed, juxtacellular or intracellular nano-stimulation of a single neuron significantly disrupts its functional integrity, can change the state of the related network, modify global brain state and even modulate behavior. The excitatory and inhibitory effects of single-cell nano-stimulation depend on multiple neuronal factors (architectural, cellular synaptic/intrinsic properties and brain state) and on the anatomical target and settings of the electrical stimulation. So, is it necessary to stimulate a large number of nerve elements to obtain therapeutic effects? Translational basic-clinic studies are required to address this question. At the Bernstein Conference 2011 we provide an overview of cell-to-network techniques. They are nowadays not applicable in human but help us to understand how small amount of exogenous current can modulate highly-distributed neural networks and subsequently the behavior.