Introduction: Deep Brain Stimulation (DBS) of the (sub)thalamic nucleus has been shown to be an effective treatment in medication-refractory Essential Tremor (ET). However, studies have reported a DBS-induced ataxia in some patients which can onset years after the start of DBS-therapy. In the ongoing debate on the cause a functional disruption of the cerebello-thalamo-cortical network is considered. In a previous study regions linked to stimulation induced side effects in the SCN, primary motor cortex, putamen, red nucleus and the posterior parietal cortex were identified. In this study we aim to test if the proposed connectivity to these nodes is associated with side-effects and could enable patient individual risk-predictions. Materials & Methods: In this study 50 patients (27 males; age 66±10 years; disease duration: 33±19 years) with Essential Tremor and bilateral thalamic DBS were clinically examined based on videos before surgery and at a follow up at least 3 years later using the Fahn-Tolosa-Martin scale (FTM) and the first three items of the scale for the assessment and rating of ataxia (SARA). Structural connectivity was estimated using a normative connectome and seeding the individual volumes of tissue activated (VTAs). Connectivity to the cortical regions previously linked to side effects was compared to the long-term tremor improvement as well to symptoms of ataxia. Based on an ordinal logistic regression analysis the best prediction model using patients‘ connectivity to these regions was identified. Using data of 40 randomly selected patients of the cohort to train the algorithm it was then tested against the remaining 10 patients. Results: On average tremor improved by 11±10 points on the FTM-scale and the SARA-score worsened by -2±2 points. Mild ataxia occurred in 21 patients and 5 patients showed severe ataxia. The normative structural connectivity to the entirety of the nodes correlated significantly to a worsening of ataxia (p < 0.01) but not to overall tremor control (p = 0.34). The best network-model prediction was based on connectivity to nodes in the primary motor cortex and the posterior parietal cortex and was able to correctly predict and categorize side effects in 7 out of 10 patients of the test cohort. Conclusion: This study supports the association of the predescribed cortical nodes to a network tied to stimulation induced ataxia side-effects in ET-patients. The results indicate that models based on the connectivity to this network could aid in the risk assessment of patients and improve individual outcome in the future.