The occurence of pulmonary edema following cocaine usage has been attributed to many factors, including the route of administration, cardiogenic factors related to myocardial ischemia and rapid increase in pulmonary vascular pressure, hypoxia of the alveolar capillary membrane, neurogenic factors, and the effects of cocaine-associated adulterants. Pulmonary vascular endothelial permeability abnormalities have been implicated in an in vivo case. We tested whether changes in pulmonary vascular endothelial permeability related to cocaine exposure occurred independently of other variables that occur in vivo. These studies were performed on confluent monolayers of cultured bovine pulmonary microvessel endothelial cells grown on gelatinized mocroporous filters. We measured the transendothelial flux of Evans blue dye following cocaine hydrochloride(10-5M) treatment as indication of vascular endothelial permeability. Results were analyzed using the Dunnett post-hoc test. We found that the cocaine-treated group had increased endothelial monolayer permeability equal to that of the positive control H2O2 (250 uM) group (clearance rates [1/min] 1.509 +/- 0.106, p= 0.01; 1.397 +/- 0.192, p= 0.01, respectively). These in vitro studies demonstrate the direct toxicity of cocaine on vascular endothelial monolayers. This toxic effect is independent of other physiologic perturbations occurring in response to cocaine exposure.