Frost-heave of soils is an important problem in engineering practice. Thus, it became necessary to find an adequate method that would reduce the negative effect of this phenomenon on building structures. The paper presents the preliminary tests results of microsilica (MS) and nanosilica (NS) influence on the water freezing process in frost-susceptible soils. The selection of stabilising additives also allowed for the analysis of changes that result from their transition from micro to nanoscale. For the purposes of the experiment a test stand was constructed, that enables to analyse six samples at the same time, at inflow of water from the bottom and freezing direction from the top. The use of linear potentiometer displacement sensors allowed to measure the increase in the height of samples automatically. Additionally, the measurement stand was equipped with digital semiconductor temperature sensors, so temperature could be measured at various heights of the samples. Moreover, an array of electrodes was applied to measure the electrical conductivity of soil at various levels of samples, which enabled the analysis of changes that occur in freezing soil. The tests were conducted according to the BS 812-124:1989 standard, in three variants: soil, soil + 5% MS, soil + 5% NS. The obtained results demonstrated that the addition of microsilica reduces the occurrence of ice lenses, which in turn significantly decreases the growth in sample height. However, the addition of nanosilica completely stops the frost-heave process. Ice lens is not created, only a frost line is noticeable. Analysing the distribution of temperatures at different heights, it was noted that stabilised samples are characterised by higher temperature in comparison to the soil without additives. This means that the analysed stabilisers reduce the degree of freezing of the soil mixture. The highest temperatures were noted in samples stabilised with nanosilica. Conductivity tests also demonstrated that the smallest changes occurred in samples stabilised with nanosilica. This results from the fact that ice lenses did not occur in such samples. In order to analyse the obtained results and their interpretation thoroughly, it is necessary to perform a detailed analysis of the microstructure both of soil and of soil with stabilising additives.