Objective: Ultrasound-guided fine-needle aspiration (UG-FNA) is utilized to sample axillary lymph nodes in breast cancer patients. Diagnostic sensitivity is good but few data exist regarding the causes of false-negative results. Study Design: Fifty-four UG-FNAs of sentinel lymph nodes with histologic follow-up were identified. Gross and radiographic lymph node size, the percentage replaced by carcinoma and the cortical thickness were correlated with false-negative rates. Results: Thirty-seven aspirates were negative, 5 of these being false-negative (9%). True-positive lymph nodes averaged 1.3 cm in dimension while false-negatives averaged 0.92 cm. Percentage involvement by carcinoma for true-positive FNAs averaged 69% while false-negatives averaged 25%. Cortical thickness averaged 5.6 mm in true-positive FNAs but 2.9 mm in false-negatives. Conclusion: A relationship exists between lymph node size and the likelihood of a false-negative FNA. Lymph nodes <1.2 cm have a higher incidence of false-negative results. Lymph nodes with <30% involvement demonstrated a higher percentage of false-negatives than those with >30% replacement. Sentinel lymph nodes <1 cm appear to be relatively poor candidates for UG-FNA. Lymph nodes with a cortical thickness <3.5 mm are more often associated with a false-negative result than nodes with a thicker cortex.