Caching has emerged as a vital tool in modern communication systems for reducing peak data rates by allowing popular fles to be pre-fetched and stored locally at end users' devices. In this paper, the concept of information theoretic security for content delivery with decentralized caching is introduced. The proposed secure caching scheme allows users to cache content in a decentralized manner and reduces the peak rate by leveraging local and global caching gains, while securely delivering requested content to the users. The analysis of the secure decentralized caching problem is presented which shows that the proposed scheme introduces security at a negligible cost in terms of achievable rate when compared to insecure decentralized caching schemes, particularly for large number of fles and users. It is also shown that the rate of the proposed secure scheme is within a constant multiplicative factor from the information theoretic optimal rate for feasible values of problem parameters.