Lymphocytes from patients with chronic active hepatitis have been found to be cytotoxic for isolated rabbit hepatocytes. Although this reaction has been shown to be of the antibody-dependent type, no autologous serum was added to the assay system and in the present experiments the source of the antibody has been sought. The failure of puromycin to block the reaction argued against a role for antibody synthesised during the culture period, and the demonstration that normal mononuclear cells could become cytotoxic when preincubated in chronic active hepatitis sera was more in favour of the passive acquisition of an antibody from the circulation. Evidence for the existence of free antibody in the sera reacting with hepatocyte surface antigens came from an additional series of experiments in which it was shown that preincubation of hepatocytes in chronic active hepatitis sera rendered them susceptible to damage by normal mononuclear cells. This effect was almost completely abolished by adding a membrane lipoportein fraction of human liver (LSP) to the sera during the preincubation step, suggesting that the LSP contained those antigens on the hepatocyte surface against which the majority of the antibodies were directed.