Abstract <p>Robot-assisted total knee arthroplasty (rTKA) has been proposed to improve precision in implant positioning and joint line restoration compared with manual TKA (mTKA). However, it remains unclear whether this increased accuracy results in superior functional or radiographic outcomes. This study aimed to compare mechanical alignment and patellofemoral restoration between robotic and manual techniques. A retrospective study including 600 consecutives primary TKAs performed from 2015 to 2024 was conducted. Patients were allocated into robotic (<i>n</i> = 300) and manual (<i>n</i> = 300) groups. All procedures were performed by the same arthroplasty team using a standardized surgical protocol and identical prosthesis model. Radiographic assessment included femorotibial mechanical axis and patellar orientation at a minimum of 12 months. Alignment was considered adequate when the absolute angular discrepancy was ≤ 2°. Statistical analyses included χ² tests, Student’s t-tests, and equivalence testing with the Two One-Sided Tests (TOST) method. Baseline demographics were comparable between groups. All knees achieved alignment within the ≤ 2° tolerance (100% vs. 100%). Mean angular discrepancy showed no significant differences (manual 0.52° vs. robotic 0.48°; <i>p</i> = 0.37). The robotic technique achieved a higher rate of exact matches (64% vs. 52%), while both remained within clinically accepted limits. Equivalence and non-inferiority analyses confirmed statistically and clinically equivalent outcomes between techniques. Robot-assisted TKA provides greater geometric precision; however, this advantage does not translate into measurable clinical or radiographic superiority when the manual technique is performed under a standardized protocol by experienced surgeons. Both methods demonstrated equivalent performance in restoring mechanical alignment and patellofemoral relationships.</p> Level of evidence <p>III. Retrospective comparative study.</p>