ABSTRACT Objective To discuss the value of gray-level ratio in differentiating between malignant and benign breast lesions. Methods The research included 328 patients（337 nodules） with biopsy in our department. Among all the nodules，183 were benign and 154 were malignant. The gray-level ratio had been calculated using the histogram of the Adobe Photoshop software. Result The mean value of the malignant lesions（0.30±0.16） was statistically lower than that of the benign lesions（0.47±0.18，）（p<0.05）. According to the ROC curve, the ratio cutoff value differentiating malignant from benign lesions was estimated to be 0.335. At this cutoff value, the ratio can predict malignancy with a sensitivity of 77.6%，specificity of 63.6%，yielding an area under the curve of 0.777. Conclusion Gray-level ratio has the potential to aid in the classification of benign and malignant breast lesions.