Recent studies suggest that the ratio of plasma alpha-amino-n-butyric acid to leucine is increased specifically by chronic heavy alcohol consumption. To test this hypothesis, we determined this ratio in normal controls and (1) currently drinking chronic heavy alcoholics; (2) currently abstaining chronic alcoholics; (3) patients with nonalcoholic liver disease; (4) chronically alcohol-fed rats and mice; (5) mice infected with murine hepatitis virus; and (6) mice exposed to carbon tetrachloride. Mean ratios in control persons, drinking alcoholics, abstaining alcoholics, and nonalcoholic liver disease patients were not statistically different. Of 5 drinking alcoholics followed serially from the beginning of abstinence, 1 had an elevated ratio on admission and a persistently elevated ratio even 2 weeks later; another had a ratio more elevated after 2 weeks of abstinence, than on admission; and 3 had ratios in the normal elevated after 2 weeks of abstinence than on admission; and 3 had ratios in the normal range. Compared to control rats, chronically alcohol-fed rats had a significantly elevated mean ratio after 1 month and an even higher ratio after 2 months (P less than 0.001). In control, alcohol-fed and carbon tetrachloride-treated mice, alpha-amino-n-butyric acid was undetectable, but in two sets of mice with severe murine hepatitis virus infection, elevated ratios were found. We conclude that because the ratio of alpha-amino-n-butyric acid to leucine is not necessarily elevated in chronic heavy alcoholics but can be elevated in acute experimental liver cell injury, it does not appear to be a specific marker for the detection of alcoholism.