Malignant and benign breast tumors present differently in their shape and size on sonography. Morphological information provided by tumor contours are important in clinical diagnosis. However, ultrasound images contain noises and tissue texture; clinical diagnosis thus highly depends on the experience of physicians. The manual way to sketch three-dimensional (3D) contours of breast tumor is a time-consuming and complicate task. If automatic contouring could provide a precise breast tumor contour that might assist physicians in making an accurate diagnosis. This study presents an efficient method for automatically contouring breast tumors in 3D sonography. The proposed method utilizes an efficient segmentation procedure, i.e. level-set method (LSM), to automatic detect contours of breast tumors. This study evaluates 20 cases comprising ten benign and ten malignant tumors. The results of computer simulation reveal that the proposed 3D segmentation method provides robust contouring for breast tumor on ultrasound images. This approach consistently obtains contours similar to those obtained by manual contouring of the breast tumor and can save much of the time required to sketch precise contours.