Due to the Leidenfrost phenomenon, immersion quenching is a more or less uncontrolled task, because prediction of the appropriate rewetting process is difficult. During the rewetting process the highest heat transfer coefficient occurs. Knowledge of the rewetting behavior is therefore one of the most important points for prediction of the distribution of heat transfer coefficients during immersion cooling. Up to now, it has been assumed that for a given set of experimental parameters, the rewetting process occurs when the surface temperature decreases below a given temperature. In the literature this temperature is called the Leidenfrost temperature. In this contribution, this assumption will be investigated by means of different kind of hollow and non-hollow cylinders. Those workpieces offer the possibility to investigate different distributions of surface temperatures and cooling rates. The observation of the rewetting process was done by means of high resolution video cameras. The investigations show clearly that for a given set of experimental parameters the rewetting process is not only a function of temperature. The progressive motion of the rewetting front is more a self-adjusting process which starts mostly on edges or surface failures of the workpiece.