This paper deals with the interaction control of robot manipulators. The ideal goal is to control the manipulator so that its end-effector reaches a desired position in the operative space. The attainment of this goal can be prevented by the fact that during its motion the end-effector can interact with the environment. So, a realistic goal is regarded as control objective in the paper, that of asymptotically reaching an equilibrium position jointly determined by the choice of the ideal desired position, and by the presence of the environment. The model of the manipulator is supposed to be affected by uncertainties, so that a classical impedance control scheme cannot be adopted. The approach proposed in this paper relies on the theory of sliding mode control. In general, this methodology is deemed nonappropriate to control mechanical systems like robot manipulators, since it introduces the so-called chattering phenomenon, which produces undesired wear and vibrations. Yet, the final control solution proposed in this paper is based on the generation of sliding modes of the second order, which practically reduces the degradation of performances, and the mechanical stress due to the chattering effect. Indeed, the proposed second-order sliding mode control scheme relies on the use of a continuous control law, in contrast to conventional sliding mode control schemes.