Consider a test result that indicates an increased risk of cancer. What if the amount of risk were not known? What if it were unclear which intervention would most effectively reduce the risk of subsequent invasive cancer? What if the result itself were subject to a great deal of interlaboratory variation? Should such a test result be used to make medical management decisions? Such tests have been in routine use in medical care for decades. For instance, the diagnosis of ductal carcinoma in situ (DCIS) of the breast is essentially the identification of a microscopic marker of increased risk of breast cancer. Even though only a minority of incompletely or untreated intraductal carcinomas progress to invasive carcinoma (1, 2), and even though the diagnosis itself is subjective and often unreliable (3), DCIS constitutes a well-established medical call to action for nearly 40,000 women each year in the U.S. The identification of cancer risk through tissue diagnosis is well-established and familiar to physicians. Through routine exposure to glass slides and photomicrographs of neoplastic cells it becomes tempting to conceptualize cancer as a disease of abnormal cellular appearance. Cancer is actually the direct result of sequential accumulation of mutations in genes required for normal growth and differentiation. Although malignancy is truly a genetic disease, the epiphenomenon of altered microscopic appearance remains to many physicians a far more comfortable basis of medical decision-making than the more abstract concept of underlying genetic mutations. Although genetic tests for hereditary cancer risk have provided valuable information to many patients and physicians, they have also been the source of controversy and consternation. A frequently stated concern is that such tests indicate an increased likelihood but not a certainty of developing cancer, and that not enough is known about the actual cancer risk conferred by mutations. It is pointed out, for instance, that while one study of women in high-risk families showed that mutations in the genes BRCA1 and BRCA2 confer a 50% risk of breast cancer by age 50 (4), another population-based study concluded that this risk was “only” 33% (5). For the patient who has a 33–50% chance of breast cancer before age 50, this range of risk would likely produce similar choices regarding health care options. Nonetheless, citing such differences in risk estimates, Dr. Bernadine Healy (6) commented that people who use genetic tests to identify hereditary susceptibility to breast and ovarian cancer are “fortunetellers . . . reading a pretty cloudy crystal ball.” In comparison, few physicians could cite the percentage of women with DCIS who develop invasive breast cancer [which in most studies is less than 50% (1)]. Similar or greater uncertainties exist for histological indicators of increased cancer risk such as carcinoma in situ, dysplasia, and atypical hyperplasia of various tissues. Every day in the practice of medicine, nonetheless, thousands of surgical interventions are prompted by such diagnoses. Physicians therefore commonly use test results that indicate an increased risk of cancer, but most such tests do so indirectly by examining cells rather than genes. Genetic tests are now available to definitively identify many hereditary cancer syndromes. The risks of cancer associated with positive results of such tests are in fact better characterized than those associated with many of the tissue-based abnormalities that physicians routinely act upon. It does not seem, therefore, that uncertainty regarding risk estimates could or should be the true basis of the distrust held by some physicians for genetic tests that identify hereditary cancer risk. Another concern expressed by some physicians is that testing for hereditary cancer susceptibility might jeopardize an asymptomatic patient’s access to health insurance. Despite (or because of) extensive discussion of this issue, it has become apparent that the perceptions of discrimination are greatly out of proportion to the reality (7, 8). The Health Insurance Portability and Accountability Act of 1996 (HIPAA) guarantees access to health insurance coverage regardless of health status and pre-existing conditions for individuals in or entering group insurance plans and specifically precludes the use of genetic information to demonstrate a pre-existing condition in otherwise healthy individuals. Although such explicit protections have not been in place for individuals undergoing other medical tests, few physicians have advised women to pay personally for Pap smears, dilatation and curettage, or breast biopsies to prevent an insurance company from learning of these procedures. There is at present no evidence that genetic testing for cancer risk in healthy individuals is accompanied by discrimination in Received February 5, 1999. Revision received March 18, 1999. Accepted March 19, 1999. Address correspondence and reprint requests to: Dr. Thomas S. Frank, Medical Director, Myriad Genetic Laboratories, 320 Wakara Way, Salt Lake City, UT 84108. 0021-972X/99/$03.00/0 Vol. 84, No. 6 The Journal of Clinical Endocrinology & Metabolism Printed in U.S.A. Copyright © 1999 by The Endocrine Society