Breast cancer is the most commonly diagnosed form of cancer in women accounting for about 30% of all cases. From a computational point of view, breast cancer diagnosis can be viewed as a pattern classification problem. In this paper, we present a cost-sensitive approach to classifying breast cancer data. In particular, we employ a fuzzy rule base that allows incorporation of a misclassification cost term in order to provide the ability to focus on certain classes and hence to boost the identification of malignant cases. Moreover, we show how genetic algorithms can be employed to optimise a compact yet effective rule base, investigating both Michigan and Pittsburgh style approaches of hybrid GA-fuzzy classifiers in the context of breast cancer diagnosis.