Objective To compare the long-term effectiveness of patellar denervation by electrotomy combined with patellar replacement and patellar denervation by electrotomy alone in primary total knee arthroplasty (TKA). Methods A retrospective analysis was conducted on the clinical data of 30 patients treated with primary TKA of both knees between July 2013 and March 2015 who met the selection criteria. There were 11 males and 19 females, aged 56-79 years, with an average age of 67.6 years. One knee was randomly selected for patellar denervation by electrotomy combined with patellar replacement during TKA (combined group), while the other knee was treated with patellar denervation by electrotomy alone (control group). All patients adopted the same type of total knee prosthesis. After surgery, the patients were followed up regularly, and the occurrence of complications was recorded. The functions of the knee and patella were evaluated using the Knee Society Score (KSS) and Feller score, respectively. The position of the prosthesis, patella trajectory, and prosthesis loosening and wear were observed by imaging examination. Results All 30 patients were followed up 81.4-103.5 months, with an average of 90.4 months. The patellar thickness of the combined group ranged from 21 to 26 mm, with an average of 23.0 mm. The position of the prosthesis and patella trajectory in the combined group and the control group were good, without obvious loosening or wear. After operation, 2 sides (6.7%) in the combined group and 3 sides (10.0%) in the control group presented joint adhesion and poor activities. No complication such as lower limb deep vein thrombosis, aseptic fractures, and infections around the prosthesis occurred in both groups. At last follow-up, the KSS clinical score, KSS function score, and Feller score showed no significant difference between the two groups (P>0.05). According to the KSS score, 24 patients (80.0%) had no obvious preference for patellar denervation combined with patellar replacement, 3 patients (10.0%) preferred patellar replacement combined with patellar denervation, and 3 patients (10.0%) preferred no patellar replacement. Anterior knee pain occurred in 6 sides (20.0%) of both groups. Conclusion There is no significant difference in the long-term effectiveness between patellar denervation combined with patellar replacement and patellar denervation alone conducted in patients with knee osteoarthritis undergoing primary TKA.