BACKGROUND: Left cervical vagus nerve stimulation (VNS) using the implanted NeuroCybernetic Prosthesis (NCP) can reduce epileptic seizures and has recently been shown to give promising results for treating therapy-resistant depression. To address a disadvantage of this state-of-the-art VNS device, the use of an alternative transcutaneous electrical nerve stimulation technique, designed for muscular stimulation, was studied. Functional magnetic resonance imaging (MRI) has been used to test non-invasively access nerve structures associated with the vagus nerve system. The results and their impact are unsatisfying due to missing brainstem activations. These activations, however, are mandatory for reasoning, higher subcortical and cortical activations of vagus nerve structures. The objective of this study was to test a new parameter setting and a novel device for performing specific (well-controlled) transcutaneous VNS (tVNS) at the inner side of the tragus. This paper shows the feasibility of these and their potential for brainstem and cerebral activations as measured by blood oxygenation level dependent functional MRI (BOLD fMRI). MATERIALS AND METHODS: In total, four healthy male adults were scanned inside a 1.5-Tesla MR scanner while undergoing tVNS at the left tragus. We ensured that our newly developed tVNS stimulator was adapted to be an MR-safe stimulation device. In the experiment, cortical and brainstem representations during tVNS were compared to a baseline. RESULTS: A positive BOLD response was detected during stimulation in brain areas associated with higher order relay nuclei of vagal afferent pathways, respectively the left locus coeruleus, the thalamus (left >> right), the left prefrontal cortex, the right and the left postcentral gyrus, the left posterior cingulated gyrus and the left insula. Deactivations were found in the right nucleus accumbens and the right cerebellar hemisphere. CONCLUSION: The method and device are feasible and appropriate for accessing cerebral vagus nerve structures, respectively. As functional patterns share features with fMRI BOLD, the effects previously studied with the NCP are discussed and new possibilities of tVNS are hypothesised.