The elevation of soluble Fas (sFas) in the sera of patients with liver disease suggests a role for sFas in the disease process; whether it is protective or not is controversial. To determine the effects of sFas on Fas-induced liver apoptosis, we manipulated mice to produce sFas by transfecting them in vivo with different amounts of an adenovirus that produces mouse sFas driven by the CMV promoter (AdsFas). Fas-mediated apoptosis was induced by administration of anti-mouse Fas (Jo2; 10 μg/mouse) one week later. The administration of AdsFas (103, 107, or 109 pfu/mouse), which was associated with only minimal side-effects, resulted in a significant reduction in the liver transaminase levels and mortality of the mice on challenge with Jo2, as compared to control mice treated with AdLacZ. However, the protective effect of AdsFas was not complete. The possibility that Jo2-induction of TNF-α in the Kupffer cells of the liver contributes to the pathology was therefore tested. Although administration of soluble TNF receptor (sTNFRI) alone did not protect the mice from the lethal effects of Jo2, administration of sTNFRI (200 μg/mouse) after infection with AdsFas (109 pfu/mouse) resulted in 100% survival of the mice on challenge with Jo2. To confirm that the production of TNF-α by Kupffer cells produce the lethal effects of Jo2 that remained after treatment with AdsFas, these cells were selectively ablated by treatment of the mice with gadolinium chloride prior to challenge with Jo2. This treatment greatly reduced early mortality and hepatocellular damage as well as TNF-α production 6 h after injection of Jo2. These results indicate that: (1) AdsFas prevents Jo2-induced apoptosis of hepatocytes; (2) In addition to mediating Fas-mediated apoptosis of hepatocytes, Jo2 can separately induce TNF-α production by Kupffer cells resulting in early mortality, and (3) Optimal protection from Jo2-induced mortality can be achieved by protection of liver cells by pretreatment with both AdsFas and sTNFRI.