OBJECTIVE The purpose of the present study was to determine the incidence of revision and report on clinical outcomes at a minimum of 10 years follow-up in patients who had received a medial unicompartmental knee arthroplasty (UKA) with an three-dimensional image-based robotic system. METHODS A total of 239 patients (247 knees), who underwent medial robotic-arm assisted (RA)-UKA at a single center between April 2011 and June 2013, were assessed. The mean age at surgery was 67.0 years (SD 8.4). Post-operatively, patients were administered the Forgotten Joint Score-12 (FJS-12) and asked about their satisfaction (from 1 to 5). Post-operative complications were recorded. Failure mechanisms, revisions and reoperations were collected. Kaplan-Meier survival curves were calculated, considering revision as the event of interest. RESULTS A total of 188 patients (196 knees) were assessed at a mean follow-up of 11.1 years (SD 0.5, range 10.0-11.9), resulting in a 79.4% follow-up rate. Seven RA-UKA underwent revision, resulting in a survivorship rate of 96.4% (CI 94.6%-99.2%). Causes of revision included aseptic loosening (2 cases), infection (1 case), post-traumatic (1 case), and unexplained pain (3 cases). The mean FJS-12 and satisfaction were 82.2 (SD 23.9) and 4.4 (SD 0.9), respectively. Majority of cases (174/196, 88.8%) attained the Patient Acceptable Symptoms State (PASS, FJS-12 > 40.63). Male subjects had a higher probability of attaining a "forgotten joint" (p < 0.001) and high satisfaction (equal to 5, p < 0.05), when compared to females. CONCLUSIONS Three-dimensional image-based RA-UKA demonstrated high implant survivorship and good-to-excellent clinical outcomes at minimum 10 years follow-up. Pain of unknown origin represented the most common reason for RA-UKA revision. METHODS III.