Objective To screen the ultrasound features of benign and malignant small breast nodules (diameter≤1cm) and to establish a logistic prediction regression model. Methods The clinical data of 215 female patients with small breast nodules were retrospectively analyzed. Pathological results were used as dependent variables, while ultrasound signs and age were used as independent variables. Binary Logistic regression based on maximum likelihood estimation in partial forward step-wise was used to establish regression model. The receiver operating characteristic curve (ROC) is used to evaluate the diagnostic efficacy of regression model. Results The internal echo uniformity, shape, margin, aspect ratio, flow grade, flow resistance index, hyperechoic halo, microcalcification, ipsilateral axillary lymphadenectasis and age were entered the Logistic equation. The different ultrasound features independently predicting diagnostic efficacy of malignant nodules in breast, which was statistically significant (P < 0.05). The correct rates of Logistic regression model and preoperative ultrasound diagnosis were 94.2% and 87.2%, respectively. The area under the curve (AUC) of logistic regression model was 0.918 (95% CI: 0.873, 0.963). Conclusions The Logistic regression model based on ultrasound signs had high prediction accuracy and high clinical applicability.