Abstract: With the tremendous growth of e-commerce use of credit cards for online purchases has increased to a great extent and it caused an explosion in the credit card fraud. Fraud has become one of the major ethical issues in the credit card industry. Fraud associated with credit card are also rising today as it has the most popular mode of payment for both online as well as regular purchase. In order to detect frauds from the mix of genuine as well as fraudulent transactions, efficient fraud detection techniques to detect them accurately are vital rather than simple pattern matching techniques. Here an approach is done to detect the credit card fraud using a hybrid approach which involve stages of pre-processing in which anonymous transactions are removed, genetic algorithm modelled for feature selection and support vector machine for classification. The proposed model is done on UCSD-FICO data mining contest 2009 dataset (anonymous and imbalanced). It is the dataset used in competition which was organized by FICO, the leading provider of analytics and decision management technology and the University of California, San Diego UCSD. This paper describes a simple fraud detection mechanism which can effectively detect fraud with great accuracy.