Lipopolysaccharide (LPS) is a component of the outer membrane of gram-negative bacteria involved in the pathogenic processes leading to postpartum metritis and endometritis in clairy cows. It also causes inflammation of the endometrium and implantation failures in many animal species. An increase in cell proliferation by LPS has been reported in human epithelial and immune cells (Martinet al. 2000 J. Immunol. 165, 139-147) but not in endometrial cells. The aim of this study was to characterise the proliferative response of bovine endometrial cells following exposure to Escheriçhia coli LPS. In vitro cultures of bovine endometrial epithelial cells (EEC) and fibroblasts were performed following collection of bovine endometrium at the slaughterhouse. Endometrial epithelial cells and fibroblasts were separated before primary culture (Charpigny et al. 1999 J, Reprod. Fertil. 117, 315-324). On passages 4 to 6, EEC issued from 5 cows (total 25 replicates) were challenged with 2, 4, 8, 12, 16, or 24 µg mL -l of LPS. At the time of challenge and 72 h later, the numbers of attached cells were counted for control and LPS-treated groups. The variation in cell numbers after the challenge was calculated as the ratio of the number of LPS-treated cells minus the number of untreated control cells to the number of untreated control cells. The variation in cell numbers overtime was analysed by ANOVA (SAS 9.1, PROC GLM; SAS Institute Inc., Cary, NC, USA) following arcsin transformation of percentages. The effect of passage number, initial number of cells at the time of challenge, treatment group, and corresponding interactions were included in the model. A significant increase in cell numbers was observed for cells treated with 2, 4, 8, and 12 µg mL -J ofLPS ( +21 ± 5%, +30 ± 5%, +41 ± 4%, and+ 14 ± 5% over the control, respectively; P≤ 0.001), whereas a nonsignificant effect was observed for 16 and 24 µg mL -t of LPS ( +4 ± 5%, and -5 ± 14%, respectively). Effects of passage number and initial number of cells at the time of challenge were nonsignificant and no interactions of those factors with treatment effects were observed. For fibroblasts, preliminary results (2 replicates from 2 cows) suggest that the response to LPS is much less important than for EEC ( +3.6 and+ 1.3% for 2 µg mL - 1 of LPS; +5.8 and +4% for 811g mL - 1 of LPS). These findings indicate that E. coli LPS stimulates proliferation of bovine endometrial epithelial cells, and that the effect of LPS is dose dependent but not linear under the range of concentrations tested.