Road salting forms an essential element in the winter maintenance of roads. With salt usage increasing there is a need for an optimizing of salt practices. This master thesis is focused on the melting of ice below 0°C with use of salt (the de-icing process). The use of dissolved salt (Sodium Chloride) is considered only, and the melting process that takes place in an adiabatic system. The increased understanding of the physics of ice melting will hopefully contribute into optimizing de-icing operations. The thermodynamics of ice melting by salt has been studied by developing a calculation model, which can be used to predict the final temperature of an “ice-salt solution” system and the amount of melted ice, for a given salt solution.Five solutions with different salt concentrations (5, 10, 15, 20 and 23 w%) have been experimentally tested to determine the temperature change inside the reactor caused by adding of a certain amount of ice to the solution. The calculated and experimentally determined values of final temperature (T_f) have been further compared to identify the discrepancy in obtained results and to which extent the developed model is theoretically applicable. The comparison between the model and experiments has shown that the model was able to predict the final temperature of the “ice-salt solution” system with high precision throughout the investigated temperature range between -16°C and 0°C. In 89% of the cases the error between the determined results lies within temperature interval ± 1°C. The model has showed to have a better performance and gives lower level of discrepancies between the calculated and measured results when the solution has reached its ice melting capacity, i.e. when unmelted ice fractions are present in salt solution. In order for decrease the level of existent uncertainties and obtain more accurate results, it is recommended to incorporate the omitted value for heat loss in the developed model and to conduct all experiments in a cold room under constant, low temperature conditions.