Investigations of the length changes upon heating and cooling due to thermal expansion alone or its combination with phase transformations are often investigated using dilatometry. In dilatometric measurements with high heating rates a large longitudinal temperature gradient within the sample hinders an accurate investigation of the physical effects. In this study we examine the longitudinal temperature gradient and its effect on the measured length change for a wide range of heating rates. Maximum temperature differences of 160 degrees C for a maximum temperature of 570 degrees C within the sample are measured. A correction method based on a numerical model of the dilatometric experiment is developed and tested. The presented method allows an accurate and flexible method for correcting the interfered length change and exact interpretation of dilatometric data for high heating rates and complex time temperature profiles. (C) 2016 Elsevier B.V. All rights reserved.