We address the importance of understanding initial states of neuronal populations and of state-dependent responses in cognitive neuroscience experiments with special emphasis on brain stimulation studies of perception and cognition. The approach we present is based on evidence that behavioural and perceptual effects of transcranial magnetic stimulation (TMS) are determined by initial neural activation state; by systematically manipulating neural activation states before application of TMS, one can selectively target specific, even spatially overlapping neural populations within the affected region. This approach is potentially of great benefit to cognitive neuroscience and remediation programmes as it combines high spatial and functional resolution with the ability to assess causality.