Despite long-term and widespread use of deep brain stimulation (DBS) in a variety of neurological conditions, the underlying mechanisms of action have been elusive. Growing evidence suggests that DBS acts through multimodal mechanisms that are not limited to inhibition and excitation of basal ganglia circuits. DBS also seems to act over variable time spans — for example, the effects on tremor are immediate, whereas the effects on dystonia emerge over several weeks — suggesting that large networks are targeted. Studies reviewing the use of DBS in pain and obsessive–compulsive disorder have demonstrated direct involvement of axonal fibres rather than grey matter. In this Review, we draw on clinical and experimental data to examine the various hypotheses that have been put forward to explain the effects of DBS. In agreement with several other experts, we suggest that the term 'deep brain stimulation' warrants modification. A potentially more accurate term is 'deep brain neuromodulation', as the mode of action spans an array of therapeutic effects over a variable period of time, and is not just limited to 'stimulation' of the basal ganglia brain centres. Terms such as 'electrical neuro-network modulation' may be useful for applications in which deep brain structures are not the primary target.