To examine the relationship of expired capnograms and respiratory system resistance (Rrs) in intubated critically ill patients, we consecutively studied 41 mechanically ventilated patients to (1) analyze the association between expired CO2 slope and auto-positive end-expiratory pressure (auto-PEEP), between Rrs and auto-PEEP, between Rrs and expired CO2 slope, and between Rrs and arterial minus end-tidal PCO2 gradient (PaCO2-PETCO2 gradient) and (2) to investigate the capacity of the expired CO2 slope and PaCO2-PETCO2 gradient to predict Rrs during mechanical ventilation. Regression analysis found a close correlation between Rrs and expired CO2 slope (r = 0.86; p < 0.001), between Rrs and auto-PEEP (r = 0.75; p < 0.001), and between auto-PEEP and expired CO2 slope (r = 0.74; p < 0.001). Weak correlation was found between Rrs and PaCO2-PETCO2 gradient (r = 0.48; p < 0.01). Prediction interval limits at 95 percent confidence level for Rrs are approximately +/- 7.39 cm H2O/L/s from the predicted value obtained by the regression equation, where Rrs = 11.42 + 2.28 expired CO2 slope. These observations suggest that CO2 elimination in critically ill patients is strongly modulated by lung, airway, endotracheal tube, and ventilator equipment resistances. Although continuous capnogram waveform monitoring at the bedside might be useful to assess Rrs, very accurate predictions could be done only in determinate patients.