This paper distinguishes between traditional screening programs which seek to detect persons with treatable diseases such as breast cancer, and recent screening programs which seek to identify and counsel persons with reducible risk factors, particularly behavioral factors such as smoking or not exercising. Screening to detect disease is based on a biomedical model, and screening for risk factors on a general susceptibility model. The two types of program can differ in goals, procedures and outcomes. An examination of these differences highlights the importance of health education in screening for behavioral risk factors, especially those which predict cardiovascular diseases. A failure to appreciate these differences has impaired the design and evaluation of risk factor screening programs.