Keeping software modules with no fault is almost impossible in a real software development. For that reason, it is important in software development to find and fix faults as early as possible. To do so, many studies have been done to analyze the relationship between the faults and software metrics. In this study, our hypothesis is that the age of software modules affects on the quality of software. To confirm this, we propose metrics that can be easily measured from the revision history. We also conducted an experiment to collect these metrics from several open source software projects, and analyze relevance to fault existence. We can successfully model the relevance between the age of modules and the rate of fault existence.