A model for foot skin temperature prediction was evaluated on the basis of 2 experiments on subjects at various environmental temperatures (light seated manual work at -10.7°C (Study 1), and a short walking period in combination with standing and sitting at +2.8°C, -11.8°C and -24.6°C (Study 2), with boots of 3 insulation levels. Insulation of the footwear was measured on a thermal foot model. Predicted and measured data showed a relatively good correlation (r=0.87) at the 2 colder conditions in Study 2. The environmental temperature of 2.8°C was not low enough at the chosen activity for a considerable foot skin temperature drop. In Study 1 the predicted temperature stayed higher for the whole exposure period and the difference between the predicted and the measured foot skin temperatures grew proportionally with time, while subsequent warm-up curves at room temperature were almost parallel. In Study 1 the correlation was 0.95. However, the paired t-test showed usually significant differences between measured and predicted foot skin temperatures. The insulation values from thermal foot measurements can be used in the model calculations. Lotens’ foot model is lacking activity as direct input parameter, however, the blood flow is used instead (effect through T_core). The Lotens foot model can give reasonable foot skin temperature values if the model limitations are considered. Due to the lack of activity level input, it will be difficult to make any good estimation of foot skin temperature during intermittent exercise. The rate of the foot temperature recovery after cold exposure was somewhat overestimated in the model - the warm-up of the feet of the subjects started later and was slower in the beginning of the warm-up than in the prediction. It could be useful to develop the model further by taking into consideration various wetness and activity levels.