Abstract A melt encasement (fluxing) technique has been used to study the effect of experimental variables, such as cooling rate, superheating time and superheating temperature, on the levels of undercooling in Cu, Cu–O and Cu–Sn melts. The maximum undercoolings achieved were 352K for pure Cu and 328 and 320K for Cu–O and Cu–Sn, respectively. The present results are higher than the ones previously reported for these particular systems. It was found that a minimum superheating temperature of 40K and a minimum superheating time of 600s are required in order to achieve undercooling of the melt prior to nucleation. However, the most efficient parameter that affected the undercoolability of the studied metal systems proved to be the number of thermal cycles applied prior to solidification. Our results clearly demonstrate the importance of the experimental parameters on the levels of undercooling of metallic melts by using a fluxing technique.