In current retail merchandise environment, electronic commerce has speedily gained a lot of attention and also provides spontaneous transactions. In electronic commerce, card payment mode has become the most important means of payment due to fast development in information technology around the world. As the usage of card payment mode increases in the last decade, magnitude of fraudulent practices is also accelerating every year. Inspite of all attempts to prevent fraud, it carries on to be a major threat to industry and government. Traditionally, organizations have concentrated on fraud prevention rather than detection, to combat fraud. In this paper, we survey number of technique for credit card fraud detection and basis on that we will propose a strong literature of Hybrid Markov Model which is a combination of Bayesian Classifier and Hidden Markov Model,  based system which will consist of the pros of all the surveyed technique and result in better fraud detection than all previous proposed system of fraud detection.