The paper analyses the loss mechanism of roof insulation kits’ performance due to dampness increase in the insulation layers. The analyzed structures were used in standard conditions for ten years and had thermal insulation made of expanded polystyrene with a hydrophobized surface. The dampness of the thermal insulation materials was determined after the referenced period for their future fitness as roof insulation, based on laboratory tests of material samples collected from the structures. They were completed with a computer simulation of heat transfer and dampness in the partition for working conditions specified for ten years, assuming the thermal conductivity was determined for the materials collected from the analyzed roofs. It was discovered that simulation-based calculation dampness values are much lower than those observed after ten years of roof utilization. Additionally, the authors attempted to determine the correlations between the period of thermal insulation materials used in real conditions and the selected properties of the products determined in laboratory tests. To that end, the collected material was dried to constant weight and then subjected to accelerated aging through total immersion in water at room temperature, for twenty-eight days, followed by 300 freeze-thaw cycles at -20ºC and +20ºC. The results helped conclude that the abovementioned laboratory testing cycle does not allow for assessing the fitness for the use of the referenced products for ten years. The directions of future laboratory tests were set, suggesting extending the testing cycle at least twice.