An unmanned surveillance robot consists of a machine gun, a laser receiver, a thermal imager, a color CCD camera, and a laser illuminator. It has two axis control systems for elevation and azimuth. A human guard can be replaced with such a robot system to take care of dangerous surveillance tasks. However, the robot system is mounded at a fixed post to take care of surveillance tasks. In order to watch blind areas, it is necessary to modify such a surveillance robot to be installed on an UGV (Unmanned Ground Vehicle) system. Thus, it is required to have a stabilization system to compensate the disturbance from the UGV. In this paper, a simulation based design scheme has been adopted to develop a mobile surveillance robot. The 3D CAD geometry model has first been produced by using Pro-Engineer. The required pan and tilt motor capacities have been analyzed using ADAMS inverse dynamics analysis. A target tracking and stabilization control algorithm of the mobile surveillance robot has been developed in order to compensate the motion of the vehicle which experiences the rough terrain. A virtual motion simulator which reproduces UGV motion has also been developed for a virtual testing of the mobile surveillance robot. To test the performance of the stabilization control system of the robot, ADAMS/simulink co-simulations has been carried out.