The purpose of our study was to assess the positive predictive value (PPV) of ultrasound (US)-guided fine-needle aspiration (FNA) and the cancer detection rate for incidentally detected abnormal axillary lymph node (LN) in patients who underwent screening US. We retrospectively reviewed 72 LNs of 69 patients (mean age, 44.9 years) who underwent US-FNA for incidentally detected abnormal axillary LNs on 50,488 screening US from January 2005 to December 2011. The PPV of US-FNA and the cancer detection rate were calculated. We evaluated US images for lymph node size, abnormal findings (hilum loss, eccentric cortical thickening, round shape, extranodal extension, or marked hypoechoic cortex), and mammography for the identification of abnormal LNs. The PPV of each finding was also calculated. The PPV of US-FNA and the cancer detection rate were 2.8% (2/72) and 0.004% (2/50,488), respectively. The mean (SD) measurements for long-axis, short-axis, and cortical thickening of the LNs were 14.9 (5.9) mm, 8.5 (3.5) mm, and 5.8 (2.8) mm, respectively. Of the positive LNs, US findings of hilum loss, eccentric cortical thickening, and extranodal extension were found, and each corresponding PPV was 6.3% (1/16), 1.8% (1/56), and 14.3% (1/7), respectively. The PPV of mammography was 14.3% (1/7). Our results suggest that the PPVs of US-FNA and the cancer detection rate for incidentally detected abnormal axillary LNs during screening US are too low to recommend axillary US during breast US screening and that follow-up is acceptable for abnormal LNs detected during screening breast US that do not have extranodal extension or are negative on mammography.