The aim of this study was to evaluate the clinical outcomes of patients undergoing off-pump robotic coronary artery bypass grafting (CABG) with either interrupted nitinol U-Clips in totally endoscopic coronary artery bypass (TECAB) or standard running suture anastomosis in robotically assisted direct coronary artery bypass (RADCAB) over a decade. From January 2007 to December 2017, 280 patients underwent robotic off-pump CABG using the da Vinci S/Si Surgical System in our centre. TECAB with interrupted nitinol U-Clips anastomosis was performed in the left internal mammary artery (LIMA) to LAD grafting in 126 patients and RADCAB (n = 154) of the LIMA to LAD was completed with standard running suture. After discharge, patients were contacted through telephone interview and were invited to attend the outpatient clinic every 6 months or 1 year. The graft patency was assessed by coronary angiography or computed tomography angiography. All cases were completed without conversion to median sternotomy or cardiopulmonary bypass. A total of 275 single internal mammary artery (IMA) grafts (271 LIMAs, 4 right internal mammary arteries) and 5 bilateral IMA grafts were used as single graft or composite grafts. All the patients were discharged without in-hospital mortality or adverse outcomes. The average follow-up was 89.7 ± 30.4 months (range, 14-143 months). The cumulative survival rates (P = 0.53), the cumulative IMA patency rates (P = 0.83), and the rates of freedom from major adverse cerebrovascular and cardiovascular events (P = 0.41) between TECAB and RADCAB all showed no significant difference in the follow-up. Robotic off-pump CABG using IMA grafts is safe and can provide reliable long-term outcomes. Compared with the standard hand-sewn running suture technique in RADCAB, interrupted suture with the nitinol U-Clips in TECAB showed similar long-term clinical results and graft patency in LIMA to LAD bypass grafting.