Road freezing caused by snowfall during wintertime causes traffic congestion and many accidents. To prevent such problems, we developed, in this study, a system to predict road freezing based on weather forecast data and the freezing generation modules. The weather forecast data were obtained from a high-resolution model with 1 km resolution for Jeju Island from 00:00 KST on December 1, 2017, to 23:00 KST on February 28, 2018. The results of the weather forecast data show that index of agreement (IOA) temperature was higher than 0.85 at all points, and that for wind speed was higher than 0.7 except in Seogwipo city. In order to evaluate the results of the freezing predictions, we used model evaluation metrics obtained from a confusion matrix. These metrics revealed that, the Imacho module showed good performance in precision and accuracy and that the Karlsson module showed good performance in specificity and FP rate. In particular, Cohen’s kappa value was shown to be excellent for both modules, demonstrating that the algorithm is reliable. The superiority of both the modules shows that the new system can prevent traffic problems related to road freezing in the Jeju area during wintertime.