Abstarct: Breast cancer mextastasis or spreading cancer beyond the breast to other organs in the body (most often the bones, lungs, liver or brain) enumerates one of the reasons for the majority of deaths from breast cancer. Detection of breast cancer mextastasis at the earliest stage helps determine the best way to contain and prediction of breast cancer progression or control of the disease and improve quality of life for patient. In this thesis, we detect the breast cancer mextastasis by using of the available information of patients records and fuzzy logic. The present study uses Mamdani inference method, Takagi-Sugeno inference method and neuro-fuzzy inference technique namely ANFIS (Adaptive Neuro-Fuzzy Inference System) to determine the breast cancer mextastasis at the early stages using patient data involved with mextastasis. The dataset used in this work is received from The Breast Cancer Research Center (BCRC). Using medical data related to actual patient, taking into account the risk factors in the prediction of mextastasis using fuzzy logic has been raised for the first time in this thesis. The strength of these methods is the use of fuzzy logic in relation to breast cancer that it's risk factors are fuzzy variables. For evaluation theses fuzzy inference systems by comparing the patient's actual output and the output from model, mean square error (RMSE) for Mamdani, Sugeno, ANFIS with the Initial FIS made by FCM and ANFIS with the Initial FIS made by subtractive clustering have been obtained 0.370894208, 0.389311797, 0.27728 and 0.31837, respectively. These values, in consultation with the physicians at the Institute for Breast Cancer as good and acceptable errors made models, were considered.