Mice of the C57BL strain, given continuous access to 10% alcohol and plain water, with unlimited food and no stress, frequently drink enough alcohol to produce intoxicating levels in the blood. Nevertheless, this behavior does not appear to replicate the essential features of human alcoholism since the drinking lacks serious toxic effects and the intoxication occurs only as transient episodes in association with homeostatic consumption of fluid and food. It is suggested that continuous monitoring of intake and estimation of the concentration of alcohol in blood, which are now technically feasible, will permit distinction between alcoholic-type drinking and a simple licking for the flavor of alcohol in beverage concentration.Consumption of alcohol by human alcoholics is both voluntary and excessive. Despite suffering from gastritis, loss of social status from repeated episodes of intoxication, and progressive deterioration of health, a severe alcoholic is likely to drink compulsively until death. The question that has occupied us and many other investigators is whether study of animals under laboratory control might shed some light on this self-destructive behavior. To date no complete counterpart of human alcoholism-life-threatening toxicity from voluntary intake-has been found in animals, although many partial models of the disease are available (1-16).The hope of understanding the basic cause of alcoholism through study of animals could be illusory since human motivation involves cognitive stresses that are not known to occur in animals. If such stress is the essential determinant of alcoholism, then animal studies will always be peripheral to the main problem, limited to analyzing the consequences rather than the causes of excessive drinking. On the other hand, this view may be too pessimistic. Some animals, such as the mice used in the present study, choose to drink solutions of alcohol in preference to plain water. Animals under constraints that force them to greater intakes of alcohol than they would exhibit if an unlimited supply of water and food were available show some of the toxic effects seen in alcoholics, but what is not known is whether these animals drink alcohol for its pharmacological effect or as a source of calories (17).The toxic effects of alcohol depend on its concentration in vulnerable tissues and on accumulation of toxic metabolites. These variables, in turn, involve a number of scale factors, such as rates of ingestion, distribution, and elimination. The differences in scale between species are substantial, and no single scale factor adequately describes the differences. The mass of a mouse is only 0.0004 times that of a man. Although the water content per unit body weight (and therefore the volume of distribution for alcohol) is about the same in the two species, the metabolism of alcohol per unit weight differs by a factor of about 5 and the rate of total oxidative metabolism per unit weight differs by a factor of about 15. Thus, the ingestion of a given quantity of alcohol p...