PURPOSE This article will examine the usefulness of diffusion tensor imaging (DTI) and diffusion-weighted imaging (DWI) on the assessment of axillary lymph nodes (ALN) of breast cancer patients. METHODS Axillary lymph nodes in 66 breast cancer patients were examined by DTI and DWI, and the largest lymph node with increased cortical thickness in axilla was selected. Morphological features, apparent diffusion coefficient (ADC), volume anisotropy, and fractional anisotropy values were measured by using a special software. Imaging findings and histopathological results were recorded. RESULTS Metastatic ALN were detected in 43 (65.1%) patients. Cortical thickness of the metastatic ALN was significantly higher than the non-metastatic ALNs (P < .001), and the long-axis-to-shortaxis ratio was significantly lower in metastatic ALNs (P < .001). There was a statistically significant difference between the ALN status and fatty hilum presence (P < .001). Apparent diffusion coefficient values of metastatic ALNs were statistically lower than those of non-metastatic ALNs (P < .001) using a cutoff value of 1.26 × 10-3 mm2 /s for b=500 ADC and 1.21 × 10-3 mm2 /s for b=800 ADC which had 97.7% sensitivity and 91.3% specificity. Fractional anisotropy and volume anisotropy values were significantly different between both groups. A cutoff value of 0.47 for b-500 fractional anisotropy had 83.7% sensitivity, 69.6% specificity 69.6% positive predictive value, and 83.7% negative predictive value. A cutoff value of 0.33 for b=500 volume anisotropy had 76.7% sensitivity, 78.3% specificity, 86.8% positive predictive value, and 64.3% negative predictive value. CONCLUSION Apparent diffusion coefficient value of metastatic ALNs was found to be significantly lower than those of non-metastatic ALN, and DTI metrics of metastatic ALN were found to be significantly higher than those of non-metastatic ALN. Overall, ADC had a better diagnostic performance than morphological features, fractional anisotropy, and volume anisotropy. Diffusion tensor imagingderived diffusion metrics may be used to complement breast magnetic resonance imaging in the future after further standardization of the imaging parameters.