OBJECTIVE: To assay peripheral inter-ictal cytokine serum levels and possible relations with non-invasive vagus nerve stimulation (nVNS) responsiveness in migraineurs. METHODS: This double-blinded, sham-controlled study enrolled 48 subjects and measured headache severity, frequency [headache days/month, number of total and mild/moderate/severe classified attacks/month], functional state [sleep, mood, body weight, migraine-associated disability] and serum levels of inflammatory markers [inter-ictal] using enzyme-linked immunoassays at baseline and after 2 months of adjunctive nVNS compared to sham stimulation and suitably matched controls. RESULTS: No significant differences were observed at baseline and after 2 months for headache severity, total attacks/month, headache days/month and functional outcome [sleep, mood, disability] between verum and sham nVNS. However, the number of severe attacks/month significantly decreased in the verum nVNS group and circulating pro-inflammatory IL-1β was elevated significantly in the sham group compared to nVNS. Levels of anti-inflammatory IL-10 were significantly higher at baseline in both groups compared to healthy controls, but not at 2 months follow-up [p < 0.05]. Concentrations of high-mobility group box-1 (HMGB-1), IL-6, tumor-necrosis factor-α (TNF-α), leptin, adiponectin, ghrelin remained unchanged [p > 0.05]. No severe device-/stimulation-related adverse events occurred. CONCLUSION: 2 months of adjunctive cervical nVNS significantly declined the number of severe attacks/month. Pro-inflammatory IL-1β plasma levels [inter-ictal] were higher in sham-treated migraine patients compared to verum nVNS. However, pro- [IL-6, HMGB-1, TNF-α, leptin] and anti-inflammatory [IL-10, adiponectin, ghrelin] mediators did not differ statistically. Profiling of neuroinflammatory circuits in migraine to predict nVNS responsiveness remains an experimental approach, which may be biased by pre-analytic variables warranting large-scale biobank-based systematic investigations [omics].