Completeness of coronary revascularization has been associated with long-term outcome [17, 21, 24]. Some evidence suggests that this may be a limit for OPCAB surgery [25], with a retrospective analyse suggesting that OPCAB is an independent predictor of incomplete revascularization [17, 20, 21, 24]. Our study showed similar completeness of revascularization between groups in the unmatched analysis but better results in the redo-CABG group in Matched Analysis A. In Matched Analysis B, completeness of revascularization favoured only slightly redo-CABG surgery (71 vs 68%; P = 0.60), although the number of grafts needed was reduced in the redo-OPCAB group (21 ± 0.8 vs 2.4 ± 0.8; P = 0.03) reflecting the baseline difference in number of diseased vessels. This finding might reflect a tendency to perform redo-OPCAB in patients with reduced number of diseased coronaries as suggested by others [22], but this speculation cannot be confirmed in our study.