ABSTRACT: BACKGROUND:Concerns about the cost and convenience of postsurgical physical therapy (PT) have sparked interest in unsupervised, home-based rehabilitation. However, the effectiveness of unsupervised home exercise after primary TKA has not been previously evaluated. QUESTIONS/PURPOSES:(1) Can unsupervised home exercise after surgery provide noninferior recovery of passive knee flexion compared with formal outpatient PT? (2) Does a web-based platform for home-based exercise provide an advantage compared with a printed PT manual? METHODS:We conducted a randomized, noninferiority trial involving 290 patients (20% of the 1464 eligible patients who could be contacted) who underwent primary TKA from March 2016 to April 2018. We included patients > 18 years old who were undergoing primary, unilateral TKA and provided written consent. We excluded patients with preoperative knee flexion < 90°, patients considering surgical intervention in a hip or the contralateral knee, patients discharged to an extended care facility, and revision or conversion TKA. We randomized patients to one of three groups: outpatient PT, unsupervised home exercise using a web-based platform (web PT), or unsupervised home exercise using a printed paper manual (paper PT). We also implemented a "delayed recovery intervention" within the home exercise program, in which patients were obliged to begin outpatient PT if knee flexion was < 70° at 2 weeks or < 90° at 4 weeks. The primary outcome was change in knee flexion from preoperative baseline after 4 to 6 weeks and 6 months. Secondary outcomes included Knee Injury and Osteoarthritis Outcome Score (KOOS), time back to activities of daily living, and time off narcotics. All analyses were intention to treat, and the noninferiority margin was 5% with maximum flexion as the outcome of interest for this parameter. RESULTS:Adjusted differences in change in passive flexion for web PT +3° (95% confidence interval [CI], -1.2° to 6.4°) and paper PT +5° (95% CI, 0.99°-8.6°) were not inferior to outpatient PT based on a predefined 5° margin. Change in knee flexion from baseline was 0° for outpatient PT, -2° for web PT, and -1° for paper PT after 4 to 6 weeks and 8°, 8°, and 12° for the three groups, respectively, after 6 months. Additionally, there was no difference in the change in KOOS from baseline at 4 to 6 weeks or 6 months postoperatively as well as time back to work, driving, and walking without an assistive device. CONCLUSIONS:Unsupervised home exercise is an effective rehabilitation strategy after primary TKA and was noninferior to formal outpatient PT in selected patients. It is worthwhile to reconsider the current practice of automatically designating patients for outpatient PT after primary TKA, because appropriately selected patients with adequate clinical support can achieve similar results at home. LEVEL OF EVIDENCE:Level I, therapeutic study.