E-banking is banking system in which the customers conduct transactions electronically via the Internet. Most of these sites are prone to a phishing attack. The usual fears with phishing attack are that the user may be an attacker and impersonating the identity of the authorized user, but the most dangerous in E-bank that the website which is to be hacker (phisher website), and this will lead to steal sensitive information of customers. The trend of this research is to introduce proposed security system represent "E-bank website anti-phishing attack". The proposal tries to prevent that the security concerns remain on the attacker impersonating an authenticateduser. In traditional authentication techniques, such as use of username and password is not sufficient for securing E-banking system. The reason of using fingerprint in authentication is based on fingerprint individuality. The proposal introduces a suggested treatment of pattern recognition of fingerprint with lowest False Acceptance Rate "FAR" and False Rejection Rate "FRR" among all traditional fingerprint authentication mechanism. The proposal enhances the traditional recognition system by applying Gabor filter during the preprocessing stage which makes our proposal able to recognize the rotated impression, so the accuracy of fingerprint matching is improved.