Carbon dioxide (CO₂) insufflation in the cardiothoracic wound cavity is used in open-heart surgery for prevention of arterial air embolism. The objective of this study was to investigate if CO₂ insufflation may influence the rate of airborne contamination of the cardiothoracic wound. This was studied in a cardiothoracic wound cavity model that contained two 9 cm blood agar plates. Contamination rates were compared between a control without insufflation and insufflation with: (1) a thin open-ended tube or a gas-diffuser, (2) air or CO₂, and (3) CO₂ flows of 5 or 10L/min. CO₂ insufflation at 5L/min with an open-ended tube resulted in a contamination rate almost four times that of the control (P=0.01), whereas with the gas-diffuser the contamination rate decreased (P=0.01). With the gas-diffuser, air insufflation at 5L/min markedly reduced the contamination rate compared with the control (P<0.001), but was less protective than CO₂ insufflation at the same flow (P<0.001). With both gases, the contamination rate was particularly low close to the gas-diffuser (P<0.001). Increasing the CO₂ flow from 5 to 10L/min reduced the average contamination rate in the model from 30% to 22% (P<0.001) of the control. At a CO₂ flow of 10L/min the contamination rate within 9 cm of the gas-diffuser was 14% of the control. Intraoperative wound ventilation with CO₂ using a gas-diffuser may not only prevent air embolism, but may also significantly reduce the risk of airborne contamination and postoperative wound infection in cardiac surgery. In contrast, insufflation with an open-ended tube substantially increases these risks.