Support Vector Machines (SVM) are supervised learning algorithms that analyze data and recognize patterns and used for classification and regression analysis. Specially, SVM is an linear classifier to perform in such a way as to minimize the generalization error. In this study, SVM apply to diagnosis breast cancer. Breast cancer test data have 9 test parameters 2 class and 683 patients. Randomly select some patients as a train data for SVM train and predict the other patient data. The diagnosis results are compared with both different train data size and different training parameters.