The soil sampling activities conducted in the plains of northern Italy adhered to the methodology outlined in the poplar inventory dataset by Rizza et al. (2023). The dataset consisted of 2,322 clusters that were georeferenced on a three-kilometre UTM grid, systematically distributed across an area spanning 4,179,600 hectares. Each cluster comprised nine subunits aligned with the points on a 500-meter square grid. The soil sampling took place between 1994 and 1996 and focused on locations with poplar plantations that were at least five years old in 1994. To obtain the samples, a pedological profile was manually excavated, and each layer was identified. One kilogram of soil was collected from each layer. Each soil sample was labelled with the region's sampling letter, cluster number, and element number. A maximum of five depths were examined for each sample. The texture characteristics (fine sand, coarse sand, fine silt, coarse silt, and clay) and certain chemical properties (limited to the first horizon) were documented, including NH4+, NO3-, CaCO3, P2O5, pH, organic matter, and soil carbon. Moreover, the presence of irrigation was recorded, and for the first layer, the depth of the temporary and permanent water table was reported. It is important to note that for certain samples, only the surface soil layers were evaluated, omitting depths exceeding 60 centimetres.