OBJECTIVE Radiofrequency thalamotomy (RF-T) is an established treatment for refractory tremor. It is unclear whether connectivity-guided targeting strategies could further augment outcomes. The aim of this study was to evaluate the efficacy and safety of MRI connectivity–guided RF-T in severe tremor. METHODS Twenty-one consecutive patients with severe tremor (14 with essential tremor [ET], 7 with Parkinson’s disease [PD]) underwent unilateral RF-T at a single institution between 2017 and 2020. Connectivity-derived thalamic segmentation was used to guide targeting. Changes in the Fahn-Tolosa-Marin Rating Scale (FTMRS) were recorded in treated and nontreated hands as well as procedure-related side effects. RESULTS Twenty-three thalamotomies were performed (with 2 patients receiving a repeated intervention). The mean postoperative assessment time point was 14.1 months. Treated-hand tremor scores improved by 63.8%, whereas nontreated-hand scores deteriorated by 10.1% (p < 0.01). Total FTMRS scores were significantly better at follow-up compared with baseline (mean 34.7 vs 51.7, p = 0.016). Baseline treated-hand tremor severity (rho = 0.786, p < 0.01) and total FTMRS score (rho = 0.64, p < 0.01) best correlated with tremor improvement. The most reported side effect was mild gait ataxia (n = 11 patients). CONCLUSIONS RF-T guided by connectivity-derived segmentation is a safe and effective option for severe tremor in both PD and ET.