Rodent models of sepsis differ from clinical human disease in that humans make substantially less whole-body nitric oxide and have different cellular responses to endotoxin. Sheep, when exposed to endotoxin, behave in a manner more similar to humans. Many studies of rodent peripheral blood mononuclear cells (PBMCs) exposed to endotoxin demonstrate increased cationic amino acid transporter function (particularly through the y+ transporter) to supply arginine substrate to upregulated nitric oxide synthase. Whether this is true in sheep is not known. We have studied cationic amino acid transport in sheep PBMCs stimulated with endotoxin, using labelled lysine. PBMCs stimulated both in vitro and in vivo show an initial reduction in total and y+ lysine transport (after 1-2 h exposure to endotoxin): a previously undescribed effect of endotoxin. In in vitro activated cells, the reduction in y+ transport was prevented by the lipoxygenase inhibitor, nordihydroguaretic acid (NDGA), and the phospholipase inhibitor 4-bromophenacyl bromide (4-BPAB), but not cyclohexamide or a number of other inhibitors of intracellular second-messenger pathways. In contrast after 14 h incubation, the expected increase in total and y+ lysine transport was seen. The increase in y+ transport could be prevented by cyclohexamide, dexamethasone, ibuprofen, the protein kinase C inhibitor sphingosine, NDGA and 4-BPAB. These results suggest that in response to endotoxin exposure there is an initial decrease in y+ activity mediated by a lipoxygenase product, followed by a substantial increase in y+ activity mediated by the products of either cyclo-oxygenase or lipoxygenase. Cyclo-oxygenase and/or lipoxygenase inhibition might be useful in reducing arginine transport, and hence nitric oxide production, in these cells. Experimental Physiology (2003) 88.2, 201-208.