OBJECTIVE We describe a simplified helium dilution technique to measure end-expiratory lung volume (EELV) in mechanically ventilated patients. We assessed both its accuracy in comparison with quantitative computerized tomography (CT) and its precision. METHODS Prospective human study. METHODS Twenty-one mechanically ventilated ALI/ARDS patients. METHODS All patients underwent a spiral CT scan of the thorax during an end-expiratory occlusion. From the CT scan we computed the gas volume of the lungs (EELVCT). Within a few minutes, a rebreathing bag, containing a known amount of helium, was connected to the endotracheal tube, and the gas mixture diluted in the patient's lungs by delivering at least ten large tidal volumes. From the final helium concentration, EELV could be calculated by a standard formula (EELVHe). RESULTS The results obtained by the two techniques showed a good correlation (EELVHe=208+0.858xEELV(CT), r=0.941; P<0.001). Bias between the two techniques was 32.5+/-202.8 ml (95% limits of agreement were -373 ml and +438 ml), with a mean absolute difference of 15%. The amount of pathological tissue did not affect the difference between the two techniques, while the amount of hyperinflated tissue did. Bias between two repeated helium EELV measurements was -24+/-83 ml (95% limits of agreement were -191 ml and +141 ml), with a mean absolute difference of 6.3%. CONCLUSIONS The proposed helium dilution technique is simple and reproducible. The negligible bias and the acceptable level of agreement support its use as a practical alternative to CT for measuring EELV in mechanically ventilated ARDS patients.