Background: Enteropathogenic and enterohaemorrhagic Escherichia coli (EPEC and EHEC respectively) infections in humans cause severe, watery diarrhea characterized by intestinal lesions. Despite much research into E. coli virulence, the effect these pathogens have on enteric ion transport is not well characterized and the underlying mechanisms have not been fully described. Methods: Monolayers of the human intestinal cell line T84 were cultured on semi-permeable membrane supports and exposed to log phase EPEC or EHEC for 2 to 8 hr (MOI=100). They were then mounted in Ussing chambers to measure basal and secretagogue-induced ion transport by monitoring short circuit current (Isc). Whole cell protein expression of the Na+-K+-Cl--cotransporter (NKCC1) was measured by Western blot in extracts of similarly infected T84monolayers.Results: EPEC, but not EHEC, increased the basal ion secretion after 4 hr exposure (control, 1.1±0.1; EPEC, 3.5±0.1; EHEC, 0.8±0.2 μA/cm2). The ∆Isc response to FSK, a cAMP-mediated secretagogue, was significantly (p<0.05) decreased after 4 hr exposure to either bacteria (control, 125.0±4.4; EPEC, 83.2±9.1; EHEC, 100.6±6.3 μA/cm2). However, the Ca2+-mediated response to carbachol was not significantly altered at any time point. Incubation with heat-killed bacteria, sterile culture medium, or basolaterally applied viable bacteria did not reproduce the altered secretory response, indicating that the effect is attachment-dependent and requires apical E. coli. Interestingly, two strains of human commensal bacteria also decreased the response to FSK after 4 or 8 hours (control, 128.1±6.5; C28/2, 4 hr, 96.1±9.0; C1, 8 hr, 82.3±9.1 μA/cm2). The secretory response to 8-Br-cAMP was significantly decreased with exposure to EPEC or EHEC (control, 125.1±2.3; EPEC, 92.4±5.8; EHEC, 105.4±7.1 μA/cm2), suggesting that the alteration is occurring downstream of cAMP-production. However, the residual FSKinduced ion transport seen with cells bathed in chloride-free buffer was not significantly decreased with infection, indicating that transport of other ions was not affected by infection. Total cellular levels of NKCC1, as shown by Western blot, were not changed by EPEC or EHEC. Inhibition of clathrin-coated pit formation using 2 μM phenylarsine oxide prior to bacterial exposure did not affect the decrease in ion secretion, suggesting that the effect was not caused by endocytosis of ion channels. Conclusion: Pathogenic and commensal E. coli decrease cAMP-dependent ion secretion of T84 cells. This effect likely occurs downstream of adenylate cyclase activity and does not involve altered NKCC1 protein expression or ion channel endocytosis.