OBJECTIVE There is some controversy as to whether there is a benefit from the use of a centrifugal pump compared with a roller pump during cardiopulmonary bypass to facilitate cardiac surgery. We compared the two pumps, with the primary aim of determining any difference in the effects on inflammation after pulmonary endarterectomy surgery which required prolonged cardiopulmonary bypass and deep hypothermic circulatory arrest. METHODS Between September 2010 and July 2013, 58 elective patients undergoing pulmonary endarterectomy were included in this prospective, randomised, controlled study; 30 patients were randomly allocated to the control group, which used a roller pump, and 28 patients to the treatment group, which used a centrifugal pump. Interleukin-6, procalcitonin, C-reactive protein, thromboelastographic parameters, P-selectin, international normalised ratio, activated prothrombin time, free haemoglobin, haematocrit, red blood cell count, white blood cell count, platelet count and protein S100β were recorded during and after the procedure. We also recorded the length of intensive care unit stay, blood loss and transfusion, neurological outcomes and respiratory and renal failure. RESULTS There was a significant difference in the primary outcome measure: Interleukin-6 was significantly higher in the roller pump group (587 ± 38 ng · l(-1) vs. 327 ± 37 ng · l(-1); p<0.001) 24 hours after surgery, which we interpreted as an increased inflammatory response. This was confirmed by a significant rise in the procalcitonin level in the roller pump group 48 hours following surgery (0.79 (0.08-25.25) ng · ml(-1) vs. 0.36 (0.02-5.83) ng · ml(-1); p<0.05). There were, however, no significant differences in clinical outcome data. CONCLUSIONS We have shown that the use of a centrifugal pump during prolonged cardiopulmonary bypass and deep hypothermic circulatory arrest is associated with a reduced inflammatory response compared to the standard roller pump. Larger multi-centre trials in this area of practice are required.