Background: The purpose of this study was to compare outcomes of patients undergoing robotic-assisted total knee arthroplasty (RA-TKA) to conventional instrumentation in the early postoperative period using traditional patient-reported outcome measures (PROMs) and wearable sensors. Methods: This was a prospective, matched, parallel cohort study of 100 patients with symptomatic end-stage knee osteoarthritis undergoing primary TKA (44 RA-TKA and 56 conventional TKA). Functional outcomes were assessed using ankle-worn inertial measurement units (IMU) and PROMs. IMU-based outcomes included impact load, impact asymmetry, maximum knee flexion angle, and bone stimulus. PROMs, including Oxford Knee Score (OKS), EQ-5D, and Forgotten Joint Score, were evaluated at pre-operative baseline, weeks 2 to 6 post-operatively, and at 3-month follow-up. Results: By post-operative week 6, RA-TKA was associated with significant improvements in maximum knee flexion angle compared to conventional TKA (118o ± 6.6o vs 113o ± 5.4o; p=0.04), symmetrical limb loading (82.3% vs 22.4%; p<0.01), cumulative impact load (146.6% vs 37%; p<0.01), and bone stimulus (25.1% vs 13.6%; p<0.01). Of note, RA-TKA demonstrated an earlier return to symmetrical limb loading, with operative limb IMU-based function reaching 80% of the non-operative limb by post-operative week 3. Significantly more RA-TKA patients achieved an 'excellent' outcome at 6 weeks compared to conventional TKA using OKS subscales (47% vs 41%, p=0.013). Conclusion: RA-TKAs were associated with functional improvements when assessed using IMUs compared to conventional TKA, which were not detected by traditional PROMs during the early post-operative period.