The relation between family history of coronary artery disease (CAD) and coronary risk variables was examined in 1,058 Utah adults. Subjects were divided into 3 family history score groups (low, intermediate and high) and the effects of 60 potential risk variables were assessed using multiple stepwise discriminant analysis. After controlling for age effects, the variables entering the discriminant function equation (p less than 0.01) were cholesterol, years smoking cigarettes, high-density lipoprotein cholesterol and triceps skinfold thickness. Although normal blood pressure variation did not discriminate among family history groups, those with a positive family history of CAD were more likely to develop hypertension than those without a family history (age-adjusted relative risk 1.79, with 95% confidence limits 1.03 and 3.09). In spite of significant differences among family history score groups with regard to risk variables, the discriminant analysis classified only 39% of subjects into the correct group. It was concluded that other unknown risk variables must mediate the effects of family history of CAD. Thus, family history should be considered an independent risk factor for CAD.