Summary Reparixin, a CXCR 1/2 antagonist, has been shown to mitigate ischaemia–reperfusion injury (IRI) in various organ systems in animals, but data in humans are scarce. The aim of this double-blinded, placebo-controlled pilot study was to evaluate the safety and efficacy of reparixin to suppress IRI and inflammation in patients undergoing on-pump coronary artery bypass grafting (CABG). Patients received either reparixin or placebo (n = 16 in each group) after induction of anaesthesia until 8 h after cardiopulmonary bypass (CPB). We compared markers of systemic and pulmonary inflammation, surrogates of myocardial IRI and clinical outcomes using Mann–Whitney U- and Fisher's exact tests. Thirty- and 90-day mortality was 0% in both groups. No side effects were observed in the treatment group. Surgical revision, pleural and pericardial effusion, infection and atrial fibrillation rates were not different between groups. Reparixin significantly reduced the proportion of neutrophil granulocytes in blood at the beginning [49%, interquartile range (IQR) = 45–57 versus 58%, IQR = 53–66, P = 0·035], end (71%, IQR = 67–76 versus 79%, IQR = 71–83, P = 0·023) and 1 h after CPB (73%, IQR = 71–75 versus 77%, IQR = 72–80, P = 0·035). Reparixin patients required a lesser positive fluid balance during surgery (2575 ml, IQR = 2027–3080 versus 3200 ml, IQR = 2928–3778, P = 0·029) and during ICU stay (2603 ml, IQR = 1023–4288 versus 4200 ml, IQR = 2313–8160, P = 0·021). Numerically, more control patients required noradrenaline ≥ 0·11 μg/kg/min (50 versus 19%, P = 0·063) and dobutamine (50 versus 25%, P = 0·14). Therefore, administration of reparixin in CABG patients appears to be feasible and safe. It concurrently attenuated postoperative granulocytosis in peripheral blood.