Transvascular transport of labeled-albumin is used to study endothelial permeability in experimental murine models of pulmonary infections. But radio-tagged albumin necessitates heavy safety procedures in terms of storage, manipulation and evacuation. The authors tested fluorescein isothiocyanate-tagged albumin (FITC-albumin) as a new marker for determination of endothelial permeability in a murine model of lung infection by Pseudomonas aeruginosa PAO1, in comparison with a standard method with (125)I-albumin. The mean permeability +/- SEM measured with (125)I-albumin was 2.45%/2 h +/- 0.37 for the control mice and 6.65%/2 h +/- 0.77 for the infected ones (P < .0001). With FITC-albumin, results obtained for both groups were respectively 4.96%/2 h +/- 0.64 and 11.5%/2 h +/- 1.2 (P < .0001). Spearman's rank coefficient was equal to .88 (P < .0001), showing a very strong correlation between both methods of measurement. The Bland-Altman analysis of bias revealed that there was no significant bias between FITC-albumin-derived and (125)I-albumin-derived values. The correction of the values obtained in plasma and lung homogenate supernatants by the subtraction of natural spontaneous fluorescence measured in these samples was crucial for the calculation of endothelial permeability in this new method. We believe that FITC-albumin can be useful for assessment of endothelial permeability in murine models of pulmonary diseases.