Abstract Transcranial direct current stimulation (tDCS) is routinely used in basic and clinical research, but its efficacy has been challenged on a methodological and statistical basis recently. The arguments against tDCS derive from insufficient understanding of how this technique interacts with brain processes physiologically. Because of its potential as a central tool in neuroscience, it is important to clarify whether and how tDCS affects neuronal activity. Here, we investigate influences of offline tDCS on network architecture measured by functional magnetic resonance imaging. Our results reveal a tDCS-induced reorganisation of a functionally-defined network that is dependent on whether we are exciting or inhibiting a node within this network, confirming in a functioning brain, and in a bias free and independent fashion that tDCS influences neuronal activity. Moreover, our results suggest that network-specific connectivity has an important role in defining the effects of tDCS and the relationship between brain states and behaviour.