Objective To investigate the effect of lateral retinacular release on the clinical outcomes after total knee arthroplasty (TKA) without resurfacing of the patella. Methods A prospective randomized controlled study was performed on 132 patients with unilateral degenerative knee arthritis undergoing TKA bewteen October 2012 and October 2014, who met the selection criteria. During TKA, lateral retinacular release was used in 66 cases (trial group) and was not used in 66 cases (control group). Two patients were excluded from the study due to missing the follow-up in trial group. Four patients were excluded from the study due to lateral retinacular release in control group. Finally, 64 patients and 62 patients were included in the trial group and in the control group. There was no significant difference in gender, age, body mass index, side, disease duration, preoperative patellar morphology, grading of patellofemoral arthritis, grade of patellar cartilage degeneration, patellar malposition, patellar maltracking, patellar score, and Knee Society Score (KSS) between 2 groups (P>0.05). The operation time, postoperative drainage volume, hospitalization time, postoperative complications, and patient satisfaction were recorded. Postoperative anterior knee pain was assessed by visual analogue scale (VAS), and the knee joint function was evaluated by KSS score and patellar score. The femoral angle, tibial angle, femoral flexion angle, and tibial posterior slope angle were measured on the X-ray film for postoperative prosthetic alignment. The postoperative patellar tracking and patellar position, as well as the presence of osteolysis, prosthesis loosening, patellar fracture and patellar necrosis were observed. Results All patients were followed up for 24 months. There was no significant difference in operation time, postoperative drainage volume, hospitalization time, and patient satisfaction between 2 groups (P>0.05). The incidence of anterior knee pain in the trial group was better than that in the control group (P=0.033). KSS score and patellar score were significantly improved in both groups at 24 months after operation when compared with preoperative scores (P<0.05), but no significant difference was found between 2 groups (P>0.05). Complications included hematoma (2 cases in the trial group, and 1 case in the control group), mild wound dehiscence (2 cases in each group respectively), skin-edge necrosis (1 case in the trial group), and superficial wound infection (1 case in each group respectively), which were cured by conservative treatment. No patellar necrosis, patella fracture, or knee lateral pain occurred in 2 groups. There was no significant difference in complication rate between groups (P=0.392). Satisfactory implant alignment was observed in both groups during follow-up. There was no significant difference in femoral angle, tibial angle, femoral flexion angle, and tibial posterior slope angle between 2 groups (P>0.05). No radiolucent line at the bone-implant interface was seen around the tibial components and femoral components in both groups. The patellar maltracking was observed in 3 patients of the trial group and 5 patients of the control group, showing no significant difference (P=0.488). However, the incidence of patellar malposition in the trial group (18.8%) was significantly lower than that in the control group (35.5%) (χ2=0.173,P=0.034). Conclusion Lateral retinacular release during primary TKA without resurfacing of the patella can reduce postoperative knee pain without increasing complications.