Inflammatory cells like eosinophils, neutrophils or mononuclear phagocytes have long been recognized as essential components in the pathophysiology of asthma. After recruitment in situ and subsequent activation, they are considered as responsible for epithelial and submucosal bronchial alterations. However, to access to the inflammatory site, these cells have to cross the endothelial wall, suggesting so a potential implication of endothelial cells (EC) in bronchial asthma. To test this hypothesis, we studied in a first step the modulation of vascular adhesions like intercellular adhesion molecule-1 (ICAM-1) on EC: supernatants of alveolar macrophages (AM) recovered by bronchoalveolar lavage in patients exhibiting a late asthmatic reaction, induced an enhanced expression of ICAM-1 on EC preparations; increase of ICAM-1 was clearly correlated to amounts of tumor necrosis factor-alpha (TNF alpha) present in AM supernatants, as shown by inhibition experiments with anti-TNF alpha antiserum. The second way to explore the possible role of EC in asthma was the detection of autoantibodies to EC in various allergic disorders: antibodies against a 120-kD EC antigen in patients with allergic granulomatosis and angiitis, antibodies towards a 55-kD component, common to human EC and platelets in patients with severe asthma, namely characterized by their corticosteroid dependence or by aspirin-induced intolerance. So our data suggest that bronchial asthma might result from either EC activation, through the induction of surface adhesion molecules or from an autoimmune process involving EC antigens.