Robotic liver resection is being introduced with its potential to overcome limitations of conventional laparoscopy. This study was undertaken to document early experience and learning curve of robotic liver resection in our institution. All patients undergoing liver resection between 2013 and 2017 were prospectively followed. Patients were divided into three consecutive tertiles (cohort I-III). Thirty-three patients underwent robotic liver resection within the study period. Twenty-four per cent of patients underwent formal right or left hemihepatectomy, 21 per cent underwent sectionectomy, 6 per cent underwent central hepatectomy, and the remainder underwent nonanatomical liver resection. Formal hemihepatectomy and right posterosuperior segment resection were undertaken in two patients in cohort I, four patients in cohort II, and four patients in cohort III. Two cases were converted to "open" operation. Operative time was 172 (194.5 ± 65.1) minutes in cohort I, 222 (247.8 ± 109.8) minutes in cohort II, and 280 (302.5 ± 84.9) minutes in cohort III, reflecting increasing degree of technical complexity. Estimated blood loss decreased significantly throughout the cohorts, being 400 mL, 200 mL, and 100 mL in cohorts I to III, respectively. Major intraoperative complications were not seen. Three patients experienced postoperative complications, resulting in a single mortality. Length of hospital stay was three days, with two patients being readmitted within 30 days. Robotic technique for liver resection is feasible and safe. It offers good short-term clinical outcomes, including for patients who require major liver resection. As the proficiency developed, a notable improvement in technically ability to undertake more complex resections with decreasing blood loss and minimal morbidity was seen.