The determination of instantaneous positions with high accuracy using differential GPS techniques is of great importance for several applications, like airborne topographic mapping, airborne gravimetry and altimetry. In some of these applications long baselines are involved creating additional problems in the correct determination of the position of the sensors. The determination of the attitude of the aircraft is also of great importance in this context. This paper presents a technique that, taking advantage of the integration of GPS and inertial measurements, produces real time positions. The demand for real time positioning introduces some more constraints, which must be taken into account from the start of the algorithm development to the end of its implementation. A methodology for processing GPS and raw data from an inertial measurement unit (IMU) was implemented in a feedback integrated structure. The position determination is based on real time DGPS with L1 and L2 carrier phase ambiguity fixing, using on-line IMU processed data to validate and increase the positions reliability, especially during GPS outages and when the aircraft is far away from the nearest GPS reference station. By processing L1 and L2 carrier phase measurements in differential mode between a pair of the aircraft GPS antennas (without resorting to ground reference station data), a partial estimate of the aircraft attitude is obtained. This is critical to align the low cost inertial platform in heading either when the aircraft is stopped or in flight. In addition, the DGPS absolute positions are used to update the IMU data in real time. By using the proposed methodology it was possible to explore the entire potential of a low cost IMU/GPS system. The position and attitude values are obtained from all the measurements through an extended Kalman filter. Decimeter accuracy was achieved for the positions. For the orientation of the aircraft an accuracy of 0.01º (1) for roll and pitch and 0.1º (1 sigma) for heading was achieved. The performance and the results obtained with this system using data from an airborne campaign that took place in Portugal are discussed.