Perimeter insulation shall limit thermal losses of the basement of buildings. Typical materials are extruded polystyrene (XPS) and cellular glass (CG). By nature of the application, mechanical stress and moisture induced loads (liquid water and diffusion), are acting on the insulation. Test standards exist for the isolated determination of compressive creep, long-term water absorption by immersion and long-term water absorption by diffusion. In real conditions, the materials are exposed to individual combinations of these parameters, depending on the depth of installation. Beside this, also system related parameters are important (i.e. sealing compound, joints), that are not considered by test methods according to the product standards. Therefore, additional requirements can be defined by building authorities. These often comprise the extraction of material from real objects. The weak point of this approach are varying conditions between different objects, limited possibilities to monitor the level of exposure and a high effort to remove the material under pressing water. This paper shows a new laboratory-based approach to assess the durability of thermal properties of a perimeter insulation. The method consists of a compression chamber, built up on top of a representative section of the insulation system installed on concrete. The compression chamber includes gravel to simulate direct contact with soil and applies a hydrostatic pressure of liquid water based on the desired depth of immersion. Temperatures are controlled on both side of the mock-up to provide a realistic diffusion gradient. Additionally, the heat flux and temperature gradient are measured. After defined time frames the material can be extracted easily and thermal conductivity, compression strength and moisture content are measured. The test scenario can be used to evaluate the fitness for purpose under controlled and realistic conditions in real time and may also be used to develop accelerated laboratory methods for the future.