In this paper, dynamics and control of a Tendon-based continuum robot is investigated. The curvature is assumed that constant in each section of continuum robot. Kinematic equation is established on the basis of the Euler-Bernoulli beam. The dynamic model of the continuum robot is derived by using Lagrange method. In this paper, robot control is performed in two parts: firstly, Dynamic model is assumed to be known and position and velocity tracking control has been by using the feedback linearization method, But uncertainties in the dynamic model, are constantly challenged the control of continuum robots. For unknown parametric quantities such as mass coefficients, one way is simply substitutes a fixed estimate for the unknown parametric quantities. In this case tracking error is not equal to zero but it’s bounded. For many applications, we cannot assume that tracking error vector is not equal to zero. In such cases we use adaptive controller. In this paper the total mass of the primary backbone and secondary backbone are uncertain parameters, therefore, a new adaptive controller is presented to estimate those uncertainties while cause to asymptotically stable for tracking error. Simulation results show good performance in velocity and position tracking.