College students received randomly assigned desirable or borderline-high cholesterol test results. Borderline-high Ss rated high cholesterol as a less serious threat to health, viewed the test as less accurate, and perceived high cholesterol as more common than did those given desirable readings. High self-esteem or blunting coping style predicted lower distress after borderline-high readings but did not moderate cognitive appraisal. In Study 2, minimization was again observed among participants in a community screening, primarily among those who had never been tested. Although the never-tested group was younger and less knowledgeable about cholesterol, these factors did not account for minimization. The results support the generalizability of laboratory data on risk appraisal and provide new evidence regarding the nature and determinants of threat minimization.