This paper presents a practical application of the principal components analysis (PCA), as one of the methods of multivariate analysis used for reducing the dimensionality of the problem (reducing a set of a large number of variables down to a smaller number of mutually uncorrelated linear combinations). Seventeen soil properties from 95 samples were analyzed in order to single out the soil properties which have significant impact on soil variability. The paper also presents a comparison of certain results obtained by using this method and by the application of factor analysis. This research showed that heavy metals contribute most to the first principal component, organic matter contributes most to the second, while inorganic compounds (oxides) contribute most to the third component. Accordingly, it is these variables that contribute most also to accountability of the total variability of the observed phenomenon. The determined accountability of the total variability by the first principal component was around 35%, by the second around 20%, while with the third it was 14%, which means that the first three dimensions together account for around 70% of the total variability of the observed set of variables.