Nasal polyps and turbinates were obtained from individuals undergoing surgery for symptomatic nasal obstruction caused by nonatopic rhinosinusitis or allergic rhinosinusitis. One part of the tissue from each patient was fixed in neutral buffered formalin and prepared for study by histochemical and immunohistochemical methods. Monoclonal antibodies were used to identify macrophages, lymphocytes, and plasma cells. In most cases (12 of 16, 75%) the remainder of the polyp and turbinate samples was treated with protease to achieve disaggregation of the epithelial cells. Those cells were cultured on permeable collagen matrix supports. Transepithelial potential difference and resistance were measured daily. At the time of maximal transepithelial potential difference, the epithelial cells were mounted in modified Ussing chambers and exposed to a sodium-positive channel blocker (amiloride hydrochloride) and to selected chloride-negative channel agonists (isoproterenol bitartrate and adenosine triphosphate). Middle turbinates and polyps were found to have more macrophages, lymphocytes, plasma cells, HLA-DR–positive cells, and eosinophils than the inferior turbinates. Epithelial cells obtained from polyps exhibited higher transepithelial potential differences and equivalent short-circuit currents than turbinate cell cultures. The responses to amiloride, isoproterenol, and adenosine triphosphate were also greater for polyp than for turbinate cultures. A theory for the pathogenesis of nasal polyps is proposed. Local release of inflammatory mediators could cause sodium absorption and chloride permeability to be higher in polyps than in turbinate epithelia. Increased sodium absorption is consistent with the hypothesis that epithelial fluid absorption contributes to the development of nasal polyps and is a result of the increased recruitment of inflammatory cells, which are present in nasal polyps. (J Allergy Clin Immunol 1997;99:165-75.)