The objective of this work is to find limitations of using MEMS-type accelerometers and gyroscopes for finding position of a vehicle in short time prospective (few seconds). Such a system could be helpful for automated vehicle driving in some situations, like bypassing a suddenly appearing obstacle. A Monte-Carlo analysis was performed to find introduced position and direction errors for various trajectories of the vehicle. Transducer noise, offset and calibration errors were taken into account as possible error sources. Also, the influence of limited data sampling rate was checked. The results are presented in form of difference statistics between real and calculated vehicle position and orientation at the end of the track as function of various parameters, including the trajectory and performance of the used sensors. The Monte-Carlo simulation accuracy was checked by bootstrap method and the errors were shown in resulting plots. Presented results show that an inertial system can be used for determining the vehicle’s position with accuracy reaching centimeters. Also, it is shown that compensation of the sensor offsets, as well as knowing the initial conditions are critical for the quality of the track reconstruction.