Bacterial translocation (BT), the migration of enteric bacteria to extraintestinal sites, is related to immune stimulation and haemodynamic changes in experimental cirrhosis. These changes may be highly relevant to patients with cirrhosis, where changes in the circulation cause serious complications. The optimal surrogate marker of BT in patients with cirrhosis, however, is a matter of controversy. In the first study, we investigated the relationship between markers of inflammation, haemodynamics and prognosis in 45 patients and 12 controls. We found high-sensitive C-reactive protein to be correlated to portal hypertension, a clinically relevant haemodynamic alteration, and appeared to be associated with increased mortality. To assess the consequences of BT on immunity, we developed an assay for the detection of bacterial DNA (bDNA), a novel marker of BT. Using the assay in the second study, in 38 patients with ascites, we found no association between bDNA and immunity, in contrast to some previous findings. In the final paper, exploring one possible translocation route, we hypothesized a difference in bDNA levels between the blood from the veins draining the gut on one hand and the liver on the other. Collecting samples during the insertion of a shunt between the two vessels in 28 patients, our finding did not suggest marked differences in bDNA, but conversely to expectations, suggested marked hepatic production of two markers of inflammation. The main results of the present thesis support some concepts of current thinking on cirrhosis pathophysiology, including the relationship of markers of inflammation to haemodynamics, disease stage and prognosis. Our results also add to a growing body of evidence suggesting that bDNA is not a clinically relevant marker of BT.