We present a system which first builds and tracks a model of a robota??s hand using a depth camera, and then uses this ability to construct a kinematic model of its own arm using very little prior information. The system is flexible, and easy to integrate with different robots, because the model building process does not require any fiducial markers to be attached to the robota??s hand. To validate the models built by the system we perform a number of experiments. The results of the experiments demonstrate that the hand model built by the system can be tracked with a precision in the order of 1mm, and that the kinematic model is accurate enough for reliably positioning the hand of the robot in camera space.