Since the mortality rate of breast cancer in women is gradually increasing, a well-designed computer-aided diagnosis (CAD) system can assist doctors in early diagnosis of the breast cancer. In this paper, a breast nodule CAD system is developed, and this system aims for a high-performance classifier for characterizing breast nodules as either benign or malignant on an ultrasonic image. A fuzzy cerebellar model neural network (FCMNN) CAD system is developed. Since the FCMNN contains the layers with overlapped membership functions, it possesses more generalization ability than a conventional fuzzy neural network. Moreover, a FCMNN can be viewed as a generation of a fuzzy neural network; if each layer of FCMNN is reduced to contain only one different neuron, then the FCMNN can be reduced to a fuzzy neural network. Thus, it is used to develop a CAD system; this is a novel research on a breast nodule ultrasound image CAD system using an FCMNN. The testing of 65 practical ultrasound images demonstrates that the proposed FCMNN CAD system can distinguish benign or malignant breast nodules with relatively high accuracy (more than 90%), and the intensive experimental results where the resulting classifier outperforms other classifiers, such as a support vector machine and a neural network by using the N -folds cross-validation method are shown. The experimental results are even higher than doctor's diagnosis; therefore, the proposed diagnostic system can serve as an assistant system to help doctors correctly diagnose breast nodules.