Background: Increasing evidence supporting the clinical effectiveness of cooled radiofrequency ablation (RFA) therapy for genicular nerves in patients with chronic knee osteoarthritis (OA) exists. However, no study has been conducted to eliminate the potential influence of a placebo effect associated with this procedure. Therefore, we evaluated the efficacy of cooled RFA compared with a sham procedure in patients with painful knees due to OA. Methods: In this double-blind, randomized, controlled study, participants were randomly assigned to receive cooled RFA of the knee (cooled RFA group, n = 20) or a sham procedure (sham group, n = 20). The primary outcome was the proportion of successful responders at the three-month follow-up. The secondary outcomes were successful responders at one and six months; pain intensity of the knee; functional status; medication; and satisfaction at one, three, and six months after the procedures. Results: For the primary outcome, the successful responder rate was significantly higher in the cooled RFA group (76.5%) than in the sham group (33.3%) (p = 0.018). For the secondary outcome, more successful responders were observed in the cooled RFA group than in the sham group at one and six months after the procedure (p = 0.041 and 0.007, respectively). The decreased knee pain intensity was maintained throughout the six-month follow-up period in the cooled RFA group. No differences were observed in functional status, medication change, or satisfaction in both groups. Conclusions: The cooled RFA of genicular nerves offers significant pain relief and surpasses the effects attributable to a placebo.