Topsoils from eight different soil series were substituted for the local soil in three arrangements involving two depths of topsoil and either local or sand subsoils. A rotation of crops was grown with uniform annual N, P and K treatments for 24 years. A peaty soil produced the highest yields of grain, straw and grass, presumably because it released more N than the mineral soils, but an average yield of swedes. No mineral soil consistently produced higher yields of all crops and, although there were some significant differences, the variations in yield were generally relatively small. The deep topsoil consistently produced the highest yields. Topsoil with sand subsoil gave the lowest yields of grain, straw and swedes but not of grass.

All the crops on the peaty soil had higher concentrations of P in the dry matter than those from the seven mineral soils. Crops and herbage on the Laurencekirk and Foud-land soils had consistently slightly higher P concentrations than those on the other five mineral soils. Compared with the amounts applied, the crops removed considerably more K but only 45–68% of the P. There were large differences in the proportion of the P residues extractable by acetic acid from the different soil series.

The depth of topsoil was at least as important as soil type in determining yield, and the results support the use of soil depth as one of the main factors used in the Land Capability Classification for Agriculture.