This retrospective study of the largest single center experience (100 patients) with off-pump robotically assisted coronary procedures in the United Kingdom (April 2002-June 2008) aimed to rationalize patient selection, describe the technique, and determine the learning curve, technical feasibility and operative outcome of robotically assisted Atraumatic Coronary Artery Bypass (ACAB). Selected patients underwent either a robotic Totally Endoscopic Coronary Artery Bypass (12) or robotically assisted ACAB (88) using a standard Da Vinci robot with three arms. A fifth of all cases had percutaneous interventions as part of a hybrid strategy. The majority of patients were overweight men. After one hundred robotic coronary procedures, this operation is now performed as part of a routine theatre list. The mean operative and total procedure times for robotically assisted atraumatic procedures were 157 and 238 min, respectively. These measurements were significantly less in the atraumatic than the totally endoscopic group with a 34.3 and 20.6% reduction, respectively (P < 0.001; equal variance not assumed). The procedural learning curve was short and independent from internal thoracic artery harvesting. We have proven conclusively that robotically assisted ACAB is feasible, more so than the totally endoscopic procedure in this particular setting. Even in the absence of an ideal stabilizer device, this procedure causes minimal disruption to the daily operating room schedule. We have also proven that body mass index is a weak predictor of the ease of robotic internal thoracic artery harvesting and should not affect patient selection.