We tested whether the ratio of alveolar dead space to alveolar tidal volume (VD(alv)/VT(alv)) can detect states of low pulmonary blood flow (PBF) in a non-invasive way. Fifteen patients undergoing cardiovascular surgeries with cardiopulmonary bypass (CPB) were studied. CPB is a technique that excludes the lungs from the general circulation. The weaning of CPB is a model that manipulates PBF in vivo because each time blood flow through the CPB decreases, expected PBF (ePBF) increases. Patients were liberated from CPB in steps of 20 % every 2' starting from 100 % CPB (very low ePBF) to 0 % CPB (100 % ePBF). During constant ventilation, volumetric capnograms were recorded and Bohr's dead space ratio (VD(Bohr)/VT), VD(alv)/VT(alv) and the ratio of airway dead space to tidal volume (VD(aw)/VT) were calculated. Before CPB, VD(Bohr)/VT was 0.36 ± 0.05, VD(aw)/VT 0.21 ± 0.04 and VD(alv)/VT(alv) 0.18 ± 0.06 (mean ± SD). During weaning from CPB, VD(aw)/VT remained unchanged while VD(Bohr)/VT and VD(alv)/VT(alv) decreased with increasing ePBF. At CPB of 80, 60, 40 and 20 % VD(Bohr)/VT was 0.64 ± 0.06, 0.55 ± 0.06, 0.47 ± 0.05 and 0.40 ± 0.04, respectively; p < 0.001 and VD(alv)/VT(alv) 0.53 ± 0.07, 0.40 ± 0.07, 0.29 ± 0.06 and 0.25 ± 0.04, respectively; p < 0.001). After CPB, VD(Bohr)/VT and VD(alv)/VT(alv) reached values similar to baseline (0.37 ± 0.04 and 0.19 ± 0.06, respectively). At constant ventilation the alveolar component of VD(Bohr)/VT increased in proportion to the deficit in lung perfusion.