This paper describes the development and feasibility testing of a multivariate equation that uses self-report information rather than physiological measures to estimate coronary heart disease (CHD) risk in a population sample of New Brunswick adults with no reported history of heart disease. The multivariate Framingham risk prediction model, which uses a variety of selfreport and physiological measures to estimate CHD risk, was first used to calculate CHD risk in the population sample. Regression analysis was then employed to identify a linear combination of "self-reportable" variables capable of closely approximating the population risk indices derived using the Framingham model. To test its utility, the self-report equation derived from the regression analysis was applied to a small telephone survey data set drawn from a second random sample of adult New Brunswickers with no reported history of heart disease. When applied to the telephone survey data, the self-report equation yielded CHD risk estimates consistent with those from the first population sample. We concluded that the development of a self-report-based methodology for assessing the CHD risk or heart health of target populations is highly feasible. Owing to the use of self-report information, as opposed to the physiological measures employed in conventional CHD risk prediction models, a self-report-based model could significantly reduce the cost of assessing CHD risk in the target populations of community-based heart health programs. Although further research will be necessary to develop a complete self-reportbased CHD risk prediction model, the results of the present study clearly indicate that this line of research has significant potential to enhance the evaluation of heart health promotion programs.