•Intermuscular coherence can function as a biomarker for Globus Pallidus interna Deep Brain Stimulation (GPi-DBS) efficacy in dystonia.•GPi-DBS decreases intermuscular coherence in the low frequency band and beta band.•Intermuscular coherence correlates with dystonia symptom severity. Finding a non-invasive biomarker for Globus Pallidus interna Deep Brain Stimulation (GPi-DBS) efficacy. Dystonia heterogeneity leads to a wide variety of clinical response to GPi-DBS, making it hard to predict GPi-DBS efficacy for individual patients. EEG-EMG recordings of twelve dystonia patients who received bilateral GPi-DBS took place pre- and 1 year post-surgery ON and OFF stimulation, during a rest, pinch, and flexion task. Dystonia severity was assessed using the BFMDRS and TWSTRS (pre- and post-surgery ON stimulation). Intermuscular coherence (IMC) and motorcortex corticomuscular coherence (CMC) were calculated. Low frequency (4–12 Hz) and beta band (13–30 Hz) peak coherences were studied. Dystonia severity improved after 1 year GPi-DBS therapy (BFMDRS: 30%, median 7.8 (IQR 3–10), TWSTRS: 22%, median 6.8 (IQR 4–9)). 86% of IMC were above the 95% confidence limit. The highest IMC peak decreased significantly with GPi-DBS in the low frequency and beta band. Low frequency and beta band IMC correlated partly with dystonia severity and severity improvement. CMC generally were below the 95% confidence limit. Peak low frequency IMC functioned as biomarker for GPi-DBS efficacy, and partly correlated with dystonia severity. IMC can function as biomarker. Confirmation in a larger study is needed for use in clinical practice.