This work presents a new mechanism for securing mobile ad hoc domain. Some human immune features and properties have been motivated in an attempt to build a robust security system. New computational techniques have been developed based on the immune system concepts, seeking to solve several engineering problems; however, not all the previous approaches have a tangible contribution to the networking security as it was not practically implemented. In this paper, a security mechanism based on immune inspired properties and features is presented. A mobile agent system has been used to map different immune components and properties. Distributed detection, first response, second response, self recovery, adaptability, and danger theory are the hallmarks of the proposed security architecture. ©2009 IEEE.