In this paper, a roll and pitch estimation algorithm using a set of accelerometers and wireless sensor networks(S/N) is presented for use in a passenger vehicle. While an inertial measurement unit(IMU) is generally used for roll/pitch estimation, performance may be degraded in the presence of longitudinal acceleration and yaw motion. To compensate for this performance degradation, a new roll and pitch estimation algorithm is proposed that uses an accelerometer array, global positioning system(GPS) and in-vehicle networks to get information from yaw rate and roll rate sensors. Angular acceleration and roll and pitch approximation are first calculated based on vehicle kinematics. A discrete Kalman filter is then applied to estimate both roll and pitch more precisely by reducing noise from the running engine and from road disturbance. Finally, the feasibility of the proposed algorithm is shown by comparing its performance experimentally with that of an IMU in the framework of an indoor test platform as well as a test vehicle.