Rewetting a hot dry surface is the establishment of wet contact between the hot surface and a liquid at a lower temperature. Rewetting occurs after destabilizing a vapor film that exists between the hot surface and the liquid. Situations involving rewetting heat transfer are encountered in a number of postulated accidents in Canada Deuterium Uranium (CANDU) reactors, such as rewetting of a hot dry calandria tube in a critical break loss-of-coolant accident (LOCA). It is also encountered in improving metals’ mechanical properties in metallurgical industries. One of the important parameters in rewetting cooling is the rewetting delay time, which is the time interval from starting to cool the surface by the liquid to the establishment of the wet contact. Determining the rewetting delay time is very important for limiting the extent of core damage during the early stages of reactor severe accidents and is essential for predicting the period after which the coolant effectively cools an overheated core. If the rewetting delay time is relatively long, an escalation in the calandria surface temperature can occur, and if the temperature was not reduced by the establishment of the wet contact, this may lead to failure of the fuel channel. Although there is increasing interest in literature in estimating the rewetting delay time of hot flat surfaces, very limited studies exist on rewetting of curved surfaces, such as tubes. In this study, experimental tests were carried out to measure the rewetting delay time at the stagnation point of hot horizontal tubes cooled by a vertical rectangular water jet. The tubes were heated to initial temperatures between 400°C and 740°C, then rapidly cooled to the jet temperature. The two-phase flow behavior was visualized using high-speed imaging, and the moment at which the vapor film collapses was captured. In addition to studying the effect of initial surface temperature on the delay time, effects of water subcooling in the range 15°C to 80°C and jet velocity in the range 0.17 to 1.43 m/s were studied and a correlation for the delay time was developed and validated. The delay time was found to strongly increase by increasing initial surface temperature and surface curvature and by decreasing water subcooling and jet velocity. The effects of solid material and tube wall thickness were also studied.