Electron paramagnetic resonance spectroscopy (EPR) was used to study free radicals and transition metal complexes in liver tissue taken from patients with liver disease. Samples were frozen to 77K directly following biopsy to prevent deterioration. Our major aim was to compare signals from patients suffering from alcohol abuse with those from patients having liver damage not induced by alcohol. Samples were obtained from 19 chronic alcohol abusers and 7 non-alcoholic liver disease patients. Of the 19 alcoholic patients, 18 had an increased fat content, 6 had Mallory's hyaline, 12 had an acute inflammatory response, 9 had increased stainable iron and 4 had evidence of fibrosis. A signal derived from free radicals with a spectroscopic splitting factor of g = 2.0045 was found in all samples. This signal in the alcoholic patients had a mean amplitude of 2.96 cm (±1.42 SD), and in patients with non-alcoholic liver disease 2.12cm (±0.82) (p = 0.10NS), measured under identical instrument settings.The molar proportion of diene conjugated linoleic acid (DCLA), a free radical marker, in the sera of alcoholic patients was 2.68% (±1.93), but did not correlate with the free radical signals obtained by EPR spectroscopy. Also, there was no correlation between the free radical derived EPR signal and fat content, Mallory's hyaline, inflammatory infiltrate, iron or fibrosis in the liver biopsy specimens. Similarly the concentrations of aspartate transaminase, albumin, and gamma-glutamyl transferase in serum samples showed no correlations with free radical concentrations.The absence of any significant increase in the stable free radical signal in the presence of alcohol induced liver disease and the lack of correlation between the signal and either histological or serological evidence of liver damage, suggests that alcohol derived free radicals may not be involved in the pathogenesis of alcoholic liver disease.Unusually large sextet features characteristic of MN(II) complexes were observed for all liver samples. Such signals are very rare in human tissue, showing that there is a strong accumulation of Mn (II) in the liver. However, no systematic trends were observed. In some samples signals characteristic of iron-sulphur cluster units were detected, but again no correlations could be discovered.