Some fields such as the Monitoring of Civil Engineering Structures, need very sensitive acceleration and inclination sensors with a high resolution at low accelerations (typically less than 1G) and moderate frequencies (typically less than 20 Hz) or very small inclinations (typically less than 1°). The instruments currently available for such measurements are either very high precision instruments (hence very costly), used in seismology and globe physics or cheaper products lacking the sensitivity needed and having inadequate measurements ranges both in frequency and intensity. Such a situation hinders considerably the updating of construction codes in seismic zones as well as the study of the complex soil-structure interaction phenomenon in risk areas. This thesis proposes to study the utilization of magnetic levitation for designing acceleration and inclination sensors, simple and sensitive enough to meet the needs of particular applications like the monitoring of civil engineering works. Levitating the seismic mass of an inertial sensor does free its design of all the constraints created by a mechanical link between the inertial mass and the base of the instrument, which render present instruments so complex and so expensive. Two levitation techniques have been investigated and used for designing and testing two prototypes. A triaxial acceleration sensor using the Active Magnetic Bearings technique and an uniaxial inclination sensor using passive diamagnetic levitation have thus been implemented and tested. Experimental results obtained with these prototypes are not better than those obtained with instruments currently available, but they do validate the concept. As part of this research, active magnetic levitation has been investigated under an angle somewhat different from its classical application since its use for the design of sensors is subject to constraints rather different from those inherent to its use as a bearing. Furthermore, since up to now, diamagnetism has found very few applications in engineering, the present study proposes an analytical model along with design criteria that can be used as a starting point for the study and implementation of micro-mechatronic systems based on diamagnetic levitation.