Nutritional epidemiology is a new term for a branch of nutrition science that is as old as the field itself. Like all scientific approaches, nutritional epidemiology is "a logical system of thought which is in a state of ev olution" (1). While authors have proposed slightly dif fering definitions of nutritional epidemiology, they generally agree that it is the study of the role that diet plays in influencing the distribution of diseases or con ditions in humans. But whether it is the work of Lind on scurvy, Goldberger on pellagra, Keys on the effects of dietary fat and cholesterol on serum cholesterol, or the monitoring of national trends in diet and nutri tional status in the National Health and Nutrition Ex amination Surveys (NHANES), the ultimate goal is the same: the reduction in the risk of or prevention of dis ability or death associated with malnutrition in all its forms. The primary focus of public health programs is on the prevention or reduction of risk in populations rather than on the treatment of individual patients. Within this context, the role of nutritional epidemiology is the identification and monitoring of nutritional factors that are etiologically significant. Building on existing re search that has defined the functions of nutrients and their interrelationships, current work in the field of nutrition involves the development of models for pre dicting outcomes based on our expanding knowledge of nutritional biochemistry. The epidemiologie ap proach has made many valuable contributions to this modeling and will continue to do so. Nutritional epidemiology provides a framework for making judgments about the etiology of associations. The reasoning process is based on the concept that cause is the summation of and interaction of many different factors. Because any disease is a result of this web of causation it may be impossible to understand the cause of a disease. It is possible to determine what some of the causes are (2). With identification of causes, public health measures may be devised to interrupt or break the web of causation or the chain of events leading to clinically defined disease, and thus reduce the risk of disease. Probably the best known examples of nutritional ep idemiology research are in the areas of diet and cardi ovascular disease (3). Serum cholesterol has been iden tified as a major risk factor for coronary heart disease (CHD), and elevated serum cholesterol is associated with a diet high in saturated fat and cholesterol. The public health approach then is to reduce the entire serum cholesterol distribution within the population by en couraging moderation in dietary intake, thereby reduc ing the risk of CHD (3). Epidemiologists often use a two-stage process of rea soning (4). In the first stage, a statistical association is established between a characteristic and a disease or condition. In the second stage, biological inferences are derived from the pattern of statistical associations. The statistical associations are based on two types of stud ies: those in which mortality or morbidity within groups are compared and those in which individuals are stud ied to determine if those with disease share a certain characteristic more frequently than those without the disease, or if those with a certain characteristic develop the disease more frequently than those without the characteristic.