Introduction. Deep brain stimulation (DBS) can provide dramatic essential tremor (ET) relief, however no Class I evidence exists. Materials and methods. Analysis methods: I) traditional cohort analysis; II) N-of-1 single patient randomised control trial and III) signal-to-noise (S/N) analysis. 20 DBS electrodes in ET patients were switched on and off for 3-min periods. Six pairs of on and off periods in each case, with the pair order determined randomly. Tremor severity was quantified with tremor evaluator and patient was blinded to stimulation. Patients also stated whether they perceived the stimulation to be on after each trial. Results. I) Mean end-of-trial tremor severity 0.84 out of 10 on, 6.62 Off, t = - 13.218, p < 0·0005. II) N-of-1: 60% of cases had 12 correct perceptions (p = 0·001), 20% had 11 correct perceptions (p = 0·013). III) S/N: > 80% tremor reduction occurred in 99/114 'On' trials (87%), and 3/114 'Off' trials (3%). S/N ratio for 80% improvement with DBS versus spontaneous improvement was 487,757-to-1. Conclusions. DBS treatment effect on ET is too large for bias to be a plausible explanation. Formal N-of-1 trial design, and S/N ratio method for presenting results, allows this to be demonstrated convincingly where conventional randomised controlled trials are not possible. Classification of evidence. This study is the first to provide Class I evidence for the efficacy of DBS for ET.