In this thesis, dynamic modeling and control of antagonistic tendon driven robotwith elastic tendons is studied in detail. In the dynamic modelling, it is assumedthat the dominant dynamics of tendons which are the longitudinal vibrations ofthe tendons can be approximated by linear axial spring model. Moreover thedynamic model of the ATDM is converted to the standard form of Singular perturbation,this allows the controller design on the base of singular perturbation theory. Theproposed control scheme consists of two major parts, slow and fast sub-control.First a slow sub-controller is designed by considering ATDS with ideal rigid tendons, which is used to control the parametric and unstructured uncertainties of the ATDS. Then, this controller is extended to the ATDS with elastictendons by adding fast sub-controller, to counteract the longitudinalvibrations caused by the inevitable elasticity of the tendons. Furthermore,to ensure all the tendons remain in tension in all configuration, theproposed control strategy benefits from the concept of internal force.In the previous research works, tension due to fast dynamics is rarely investigated but our proposed control algorithm guarantees the tendonsin tension at all times. Tikhonov's theorem is used to analyze the stability of the overall closed-loop system. Finally the efficiency of the proposedcomposite control algorithm is investigated through simulations and experimentalresults. For the experimental investigation of proposed control law an ATDSis designed. This paper also addressed the applications of ATDJ and integratedversion of ATDJs. And it is demonstrated that ATDS is an appealing candidatefor Humanoid robotic Arm design.