To evaluate arterial (PaCO2), end-tidal (PETCO2) and carbon dioxide tension difference during mechanical ventilation with extratracheal airways, 60 patients ASA physical status I-II, receiving general anaesthesia for minor extra-abdominal procedures were randomly allocated to receive either a cuffed oropharyngeal airway (group COPA, n=30) or a laryngeal mask (group LMA, n=30). The lungs were mechanically ventilated by IPPV using a 60% nitrous oxide and 1–1.5% isoflurane in oxygen mixture (VT8 mL kg−1; RR=12 b min−1; I/E=1/2). After PETCO2 had been stable for at least 10 min after airway placement, haemodynamic variables and PETCO2 were recorded and an arterial blood sample was obtained for measurement of PaCO2. No differences in anthropometric parameters, smoking habit, haemodynamic variables and incidence of untoward events were observed between the two groups. Airway manipulation, to maintain adequate ventilation, was required in only nine patients in the cuffed oropharyngeal airway group (30%)(P < 0.0005); however,in no case was it necessary to remove the designated extratracheal airway due to unsuccessful mechanical ventilation. The mean difference between arterial and end-tidal carbon dioxide partial pressure was 0.4±0.3 KPa in the laryngeal mask group (95% confidence intervals: 0.3–0.5 KPa) and 0.3±0.26 KPa in the cuffed oropharyngeal airway group (95% confidence intervals: 0.24–0.4 KPa) (P =NS). We conclude that in healthy adults who are mechanically ventilated via the cuffed oropharyngeal airway, the end-tidal carbon dioxide determination is as accurate an indicator of PaCO2 as that measured via the laryngeal mask, allowing capnometry to be reliably used to evaluate the adequacy of ventilation.