Haptic teleoperation is a promising approach for dealing with the manipulation of micro-objects, fabricated in small series or as prototypes, and in processes which are novel or uncertain. Human operators provide their ability to plan, understand and react when faced with unexpected situations during the manipulation process, while robotic devices ensure the levels of precision required by the tasks. In order to improve the state of this field, this thesis intends to understand how to better support a human operator performing micromanipulation tasks, and based on that understanding develop a system for teloperated micromanipulation, focusing on the slave-side devices. The first stage of this research is an in-depth analysis of the requirements coming from the kind of tasks which the system must deal with, from the support that is possible and useful to give to the human operator, and from the abilities and limitations of that user. Following those requirements, a system-concept is developed, consisting of the integration of commercially available products with custom developed components. In particular, a 6 DOF magnetic levitation stage is developed as a fine positioning stage for the slave robot, achieving a movement range of 200 x 200 x 200 micrometers and rotations of 18 to 42 mrad, with Minimum Incremental Motion of 50 nm and 3.5 to 7 microradians. A silicon based force sensor is also developed to explore possibilities for force and torque sensing during micromanipulation. This force sensor measure loads in 6 DOF, within a range of 4 to 30 mN in forces and 4 to 50 uNm in torques, and with noise levels up to 13 to 27 uN/sqrt(Hz) and 11 to 43 nNm/sqrt(Hz). The system is integrated and characterized, and its usefulness is demonstrated through the performance of micromanipulation tasks by human operators. A general conclusion drawn from this research is that in order to make haptic teleoperated micromanipulation systems a viable and competitive option, it is vital to identify the kind of tasks for which haptic teleoperated micromanipulation systems can be a solution, and to optimize such systems and its components for these applications and for the haptic teleoperation scenario. In order to do so, one must understand both the advantages and limitations that this approach offers compared to its main competitors: automatic manipulation, self assembly, unaided manipulation by hand, among others. In particular, the highest potential of teloperated systems is on dealing with uncertain situations, thanks to the reasoning abilities of the human operators. Therefore, the use of these systems in structured and repetitive tasks does not constitute a fair demonstration of their advantages. Likewise, the use of components meant for automatic manipulation which often over-perform some of the motor abilities of the user, results in systems which are more complex and expensive than required, thus undermining some of the main advantages of using teleoperated systems. Following that reasoning, this work places particular attention to the definition of the requirements. By carefully studying the consequences of including a human operator in the system, and the special needs arising from the tasks and support modes, it is possible to optimize system components for this particular niche. Thus, the resulting system can deal with the situations normally encountered in teleoperated micromanipulation, without incurring in significant costs or complexities often found in systems intended for automatic manipulation, and without having to compromise properties useful for this application.