Abstract In the frame of investigating the fundamental nature of gravity, the Laser Interferometer Space Antenna (LISA) mission could open the way to a new kind of observations unreachable from ground. The experiment, based on a V-formation of six drag-free spacecraft, uses the cubic proof masses of inertial sensors to reflect the laser light, acting as reference mirrors of a 5 × 109 m arm length interferometer. The proof masses are also used as inertial references for the drag-free control of the spacecraft which constitute in return a shield against external forces. Derived from space electrostatic accelerometers developed at ONERA, such as GRADIO for the ESA ARISTOTELES and now GOCE mission (Bernard and Touboul, 1991) , the proposed LISA sensor should shield its proof mass from any accelerometric disturbance at a level of 10 −15 ms −2 Hz − 1 2 . The accurate capacitive sensing of the mass provides its position relative to the satellite with a resolution better than 10 −9 m Hz − 1 2 in order to control the satellite orbit and to minimise the disturbances induced by the satellite self gravity or by the proof mass charge. The sensor configuration and accomodation has to be specifically optimised for the mission requirements. Fortunately, the sensor will benefit from the thermal stability of the LISA optical bench environment, i.e. 10 −6 K Hz − 1 2 , and of the selected materials that exhibit a very low coefficient of thermal expansion (CTE), ensuring a high geometrical stability. Apart from the modeling and the evaluation of the flight characteristics, the necessary indirect ground demonstration of the performance and the interfaces with the drag-free control will have to be considered in detail in the future.