In recent years, great emphasis has been placed on the introduction of energy-saving solutions to the construction sector. Building envelopes made of concrete with a specially selected composition give great opportunities in this regard. As part of a wide-ranging experiment, the authors undertook to diagnose how much thermal conductivity, volumetric specific heat and thermal diffusivity can be improved with an aerating admixture and different types of aggregates. Three groups of composites were tested: B1 – on stone aggregate, B2 – on expanded clay aggregate, B3 – on sintered fly ash aggregate. Each of the groups was divided into 4 formulations made without an aerating admixture and with its increasingly higher content of 0.8, 1.1, 1.4% in relation to the weight of cement. The thermal parameters were measured on the top (T) and bottom (B) surfaces of 36 rectangular samples (3 samples from each of the 12 mixtures) with the ISOMET 2104 apparatus. Diagnostic tests concerning the influence of measurement conditions were carried out on dry and water-saturated samples. It has been proven that for each composite and in both conditions, the values of thermal parameters determined on the lower surfaces will not correctly describe the properties of the real structure present in the main volume of the element. Only measurements carried out on surfaces with a structure corresponding to the interior of the element provide adequate data that can be used in decision-making processes and in numerical simulations to assess the real thermal qualities of building envelopes.