Purpose: Misdiagnosed breast cancer on screening mammography was retrospectively reviewed in an attempt to analyze the cause for the undiscovery and to determine findings that might have been suggested on the mammographic examination. Materials and Methods: Of the 173 breast cancers that were pathologically proven, 17 breast cancers misdiagnosed on mammogram was retrogradly analyzed for the causes of the misinterpretation and predominant secondary findings. Results: Three of the 17 were negative on mammography except for dense breast. Two patients had localized asymmetricity without visible mass. Ten patients had benign-looking mass which was well-defined(n=5), partly defined(n=2), multiple (n=1), or included benign-looking calcification(n=2). Remaining two cases werer etrospectively diagnosed as misinterpreted breast cancers. Various secondary findings such as were useful in the diagnosis of breast cancer. Conclusion: Meticulous observation could improve the diagnostic yield of breast cancerin mammography.