Objectives: To determine a) whether hydrogen peroxide-induced, early lung endothelial cell dysfunction can be detected in an isolated, perfused, rat lung model; and b) whether the organic phosphothioate N-(2-mercaptoethyl)-1,3-propanediamine, which protects cells in culture against hydrogen peroxide-mediated damage, can exert the same protection in this model. Design: Intervention study; before-after trial. Setting: Research laboratory. Model: Isolated, perfused, rat lung model. Intervention: Continuous hydrogen peroxide infusion at increasing concentrations and infusion times; preceded or not by a N-(2-mercaptoethyl)-1,3-propanediamine infusion. Measurements and Main Results: Early pulmonary endothelial cell alterations, assessed by the lung extraction (% extraction) of I-123-metaiodobenzylguanidine. Permeability edema by % extraction of I-125-human serum albumin and the lung dry-to-wet weight ratio. Control experiments: % extraction-I-123-metaiodobenzylguanidine: 21.7 +/- 3.8% (n = 7). With increasing concentrations of hydrogen peroxide (0.025, 0.125, 0.5, and 2 mmol), % extraction-I-123-metaiodobenzylguanidine was progressively depressed (n = 28, ANOVA, p <.05), significantly decreased from controls at 2 mmol (10.2 +/- 5.0%, n = 7, p <.05). When the 2-mmol hydrogen peroxide infusion was preceded by the N-(2-mercaptoethyl)-1,3-propanediamine (2 mmol) infusion, % extraction-I-123-metaiodobenzylguanidine (19.9 +/- 2.9%, n = 5) was not significantly different from controls (n = 7) and was significantly greater than after the 2-mmol hydrogen peroxide infusion alone (8.7 +/- 7.4%, p <.05, n = 8). In all experiments, % extraction of human serum albumin ratio and dry-to-wet weight ratio were not significantly different from that of controls. Conclusions: a) Hydrogen peroxide-induced lung endothelial cell dysfunction was detected at an early stage, before any permeability defect appeared; b) N-(2-mercaptoethyl) -1,3-propanediamine protected against such damage.