The objective of this study was to determine whether assessment of stroke volume (SV) and measurement of exhaled end-tidal carbon dioxide [Formula: see text] during an end-expiratory occlusion (EEO) test can predict fluid responsiveness in the operating theatre.Forty-two subjects monitored by oesophageal Doppler who required i.v. fluids during surgery were studied. Haemodynamic variables [heart rate, non-invasive arterial pressure, SV, cardiac output (CO), respiratory variation of SV (ΔrespSV), variation of SV during EEO, and E'(CO₂) were measured at baseline, during EEO (Δ(EEO)), and after fluid expansion. Responders were defined by an increase in SV over 15% after infusion of 500 ml of crystalloid solution.Of the 42 subjects, 28 (67%) responded to fluid infusion. A cut-off of >2.3% ΔSV(EEO) predicted fluid responsiveness with an area under the receiver-operating characteristic (AUC) curve of 0.78 [95% confidence interval (95% CI): 0.63-0.89, P=0.003]. The AUC of ΔrespSV was 0.89 (95% CI: 0.76-0.97, P<0.001). With an AUC of 0.68 (95% CI: 0.51-0.81, P=0.07), E'(CO₂)(EEO) was poorly predictive of fluid responsiveness.ΔSV(EEO) and ΔE'(CO₂) were unable to accurately predict fluid responsiveness during surgery.