Background: Few randomized controlled trials with a midterm follow-up have compared matrix-assisted autologous chondrocyte transplantation (MACT) with microfracture (MFx) for knee cartilage lesions. Purpose: To compare the structural, clinical, and safety outcomes at midterm follow-up of MACT versus MFx for treating symptomatic knee cartilage lesions. Study Design: Randomized controlled trial; Level of evidence, 1. Methods: A total of 48 patients aged between 18 and 50 years, with 1- to 4-cm2 International Cartilage Repair Society (ICRS) grade III to IV knee chondral lesions, were randomized in a 1:1 ratio to the MACT and MFx treatment groups. A sequential prospective evaluation was performed using magnetic resonance imaging (MRI) T2 mapping, the MOCART (magnetic resonance observation of cartilage repair tissue) score, second-look arthroscopic surgery, patient-reported outcome measures, the responder rate (based on achieving the minimal clinically important difference for the Knee injury and Osteoarthritis Outcome Score [KOOS] pain and KOOS Sport/Recreation), adverse events, and treatment failure (defined as a reoperation because of symptoms caused by the primary defect and the detachment or absence of >50% of the repaired tissue during revision surgery). Results: Overall, 35 patients (18 MACT and 17 MFx) with a mean chondral lesion size of 1.8 ± 0.8 cm2 (range, 1-4 cm2) were followed up to a mean of 6 years postoperatively (range, 4-9 years). MACT demonstrated significantly better structural outcomes than MFx at 1 to 6 years postoperatively. At final follow-up, the MRI T2 mapping values of the repaired tissue were 37.7 ± 8.5 ms for MACT versus 46.4 ± 8.5 ms for MFx ( P = .003), while the MOCART scores were 59.4 ± 17.3 and 42.4 ± 16.3, respectively ( P = .006). More than 50% defect filling was seen in 95% of patients at 2 years and 82% at 6 years in the MACT group and in 67% at 2 years and 53% at 6 years in the MFx group. The second-look ICRS scores at 1 year were 10.7 ± 1.3 for MACT and 9.0 ± 1.8 for MFx ( P = .001). Both groups showed significant clinical improvements at 6 years postoperatively compared with their preoperative status. Significant differences favoring the MACT group were observed at 2 years on the KOOS Activities of Daily Living ( P = .043), at 4 years on all KOOS subscales (except Symptoms; P < .05) and the Tegner scale ( P = .008), and at 6 years on the Tegner scale ( P = .010). The responder rates at 6 years were 53% and 77% for MFx and MACT, respectively. There were no reported treatment failures after MACT; the failure rate was 8.3% in the MFx group. Neither group had serious adverse events related to treatment. Conclusion: Patients who underwent MACT had better structural outcomes than those who underwent MFx at 1 to 6 years postoperatively. Both groups of patients showed significant clinical improvements at final follow-up compared with their preoperative status. MACT showed superiority at 4 years for the majority of the KOOS subscales and for the Tegner scale at 4 to 6 years. The MACT group also had a higher responder rate and lower failure rate at final follow-up. Registration: NCT01947374 ( ClinicalTrials.gov identifier).