The data mining Techniques are used and applied in the various fields of science. The software mining is an application of data mining, which describes investigation into the examination of software Repositories, such as software assurance, reuse, fault, effort or cost prediction and detection of incomplete changes. Fault in software system is a deficiency that causes software failure and also affects software quality, cost and time. Software quality assurance is major vital to increase the level of the developed software. In existing system different types of rule based classification Algorithms are used for the fault prone modules, In Proposed , A new rule based classification algorithm ERA(Enhanced Ripper Algorithm) has be introduced. The ERA is the enhanced form of Ripper algorithm, in this algorithm we classify the software modules through Rule’s and it is mainly designed to minimize the Rule set length. For that we improve classification accuracy. Which include the attribute selection, to improve accuracy and software efficiency. In this research Enhanced ripper algorithm applied on publicly available datasets of NASA Repository and predicted the fault modules.