How different the practice of oncology would become if physicians could predict which of their patients will develop cancer, and, if the disease does occur, then determine who might remain disease-free. Programs stressing prevention, early detection, and prompt treatment could be armed at the approprrate populatron, relieving anxiety in those destined to remam disease-free, and targeting treatment to improve what might have been the predicted outcome. Not only would our patients benefit, but expensive resources could be allocated more effectively.