To reflect the extent of thermolesion of ganglion by testing the change of trigeminal somatosensory-evoked potential (TSEP) before and after ganglion radiofrequency thermocoagulation surgery (GRT), and evaluate long-term clinic effect by follow-up visiting of 1 year. Patients with idiopathic trigeminal neuralgia (TN) in the second division were enrolled between October 2014 and October 2015. They were treated with computed tomography-guided GRT and a follow-up visiting of 1 year. Bilateral TSEP measurements were performed 1 day before and 2 days after the GRT surgery. The latency and peak-to-peak amplitude of W2 and W3 were recorded. Immediate postprocedure pain relief (grades I-III) was 100% and 92.5% 1 year later. Facial numbness rate of grades III and IV was 70%, 40%, and 12.5%, respectively, at immediate, 2 days, and 1 year after GRT. No sever complications happened. The latency of W2 and W3 of patients who had no pain no numbness after 1 year of GRT was 1.74 +/- 0.24 and 3.84 +/- 0.66 ms, respectively, of TN side, and 1.71 +/- 0.39 and 3.63 +/- 0.85 ms of the healthy side before GRT. The amplitude of W2 and W3 was 1.13 +/- 0.50 and 1.99 +/- 1.09 uv, respectively, of TN side and 1.24 +/- 0.40 and 1.89 +/- 0.81 uv of the healthy side before GRT. There was no statistical difference of the latency and amplitude between 2 sides of W2 and W3 before surgery (P>0.05). The latency of W2 and W3 delayed and the amplitude reduced especially in TN side after surgery comparing before (P<0.001). And, comparisons of the latency and amplitude of W2 and W3 between TN side and the healthy side after surgery showed the latency of W2 and W3 delayed (W2: P=0.02; W3: P=0.01) and the amplitude of W2 reduced (P=0.003), but the amplitude of W3 had no statistical difference (P=0.22). The mean delayed latency and 95% confident interval of W2 and W3 were 0.22 +/- 0.35 (0.1-0.34) ms and 0.35 +/- 0.64 (0.14-0.57) ms, respectively. The mean decreased amplitude and 95% confident interval of W2 and W3 were 22 +/- 24 (14-30)% and 23 +/- 32 (12-34)%, respectively. GRT can make the latency delay and the amplitude decrease of TSEP. And the latency and amplitude of W2 and W3 can be considered reliable and safe reference for monitoring the extent of thermolesion.