OBJECTIVE: To explore the effectiveness of a NMES device in improving pain and function for individuals with KOA. METHOD: This was a randomised, sham-controlled, superiority trial recruiting individuals with symptomatic KOA, with both participants and outcome assessors blinded to the treatment allocation. Participants were randomised to receive either self-administered NMES, consisting of quadriceps (5 times a week) and footplate stimulation (7 times a week), or a sham device for 8 weeks. The primary outcome was the change in Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain domain at week 8. RESULTS: 179 participants were randomised; 50.8 ​% female, mean (SD) age 66.9 (8.7) years. 164 attended the follow up visit. No difference in the WOMAC pain domain was seen between groups (-0.7; 95 ​% CI: -1.5, 0.1; P ​= ​0.10), though within group improvements were reported in both groups (NMES: -1.5; 95 ​% CI: -2.0, 0.8; P ​< ​0.01, sham: -0.8; 95 ​% CI: -1.4, 0.2; P ​< ​0.01). There were no differences in strength, exercise and functional capacity, swelling, health related quality of life or anxiety and depression between groups. Sub-group analysis for NMES compliance or recruitment source did not change response in primary outcome. CONCLUSION: Use of an NMES device on the quadriceps and plantar flexors did not enhance pain relief or improve function over the sham control in people with KOA.