In this work, a numerical hand-glove-mitten model is developed to simulate human hand physiological responses in various cold environments. The model was validated against data from published literature and human trials conducted at -20, -40, and -60 °C in a chamber. Results demonstrated that the model could reasonably predict cold-induced vasodilation (CIVD) responses at 0 °C. At -20 °C, the model predicted skin temperature with the root mean square deviation (RMSD) falling within the measurement SD for both the entire/local hand except at the posterior hand. At -40 and -60 °C, the model could predict the trend of the skin temperatures of the whole/local hand, but the RMSD was greater than the SD for the majority of the predictions. A parametric study revealed that the palm and posterior hand exhibited higher skin temperatures than fingers, while the thumb had the lowest skin temperature of the fingers for all cases except for the case with a 3.5 clo mitten at -60°C. The proposed numerical hand-glove-mitten model provides fundamental knowledge for cold stress prediction and protective gloves development, which stands to improve the safety and health of industrial workers, firefighters, first responders, and troops.