BACKGROUND: The focus of teaching for clinical breast evaluation has been the technique of breast examination. This study questions the relationship between breast examination technique and the ability to detect physical findings. METHODS: This study examines the relationship between breast examination skills of 66 graduating primary care physicians as measured during an objective structured clinical examination (OSCE) and lump detection sensitivity and specificity on breast models. RESULTS: Overall breast examination performance revealed 50% of maneuvers performed correctly. Mean breast model sensitivity for lump detection was 40% and the mean breast model specificity was 77%. While a mild correlation existed between breast examination skills and lump detection sensitivity (r = .34, P = 0.01), no relationship was found between lump detection specificity and examination skills. CONCLUSIONS: There is a limited relationship between correct performance of breast examination maneuvers and the ability to detect a breast lump when present. Breast examination skills and palpation skills to detect masses may represent independently acquired skills with need for separate instructional methodology. These results raise serious concerns about the reliance on standardized patients alone for training in physical examination skills.