Of 489 asymptomatic women who were referred to our institute from other screening clinics, 46 were eventually proven to be breast cancer patients, and this number equated to 8.1% of the 565 breast cancer patients treated in our institute during the period of January 1997 to December 1998. Among the 46 cancer patients of the initial 489 asymptomatic women, twenty-five(54.35%) were detected by mammogram alone, six(13.04%) by clinical breast examination(CBE) alone, and the remaining 15(32.61%) by both mammogram and CBE. In context with age, the mammographic sensitivity for cancer detection was 100% for women aged over 60, 91% for 50s, 78.9% for 40s, and 75% for 30s, and inversely correlated with the patient's age. Among the 25 cancers detected by mammogram alone, 18(72%) belonged to DCIS or stage|. In contrast, four(66.7%) of six cancers detected by CBE alone and nine(60%) of 15 cancers by both CBE and mammography were included in stage II a or II b. However, the total incidence of early cancers(stage 0 and I) was significantly higher in the screening group than in the symptomatic group(P<0.01). These results suggest that the role of mammography is important in the detection of cancers in their earlier stage and CBE is helpful in reducing false negative results in breast cancer screening. In conclusion, film mammography is the best tool for the detection of microcalcification and is useful for the detection of earlier lesions, but is not perfect for the detection of breast cancer particularly in young women. A careful CBE is an essential part of breast screening in order to reduce false-negative results.