Experiments with main-crop potatoes were made at Rothamsted and Woburn in 1971 and 1972 to measure the effects of applying four amounts (1260, 1880, 2510 or 3140 kg/ha) of granular NPK fertilizer (containing 13% N, 13% P2O5 and 20% K2O)for two pairs of varieties (Pentland Crown and King Edward at Rothamsted; Pentland Crown and Record at Woburn) planted either 30 or 45 cm apart in the row.

Yields of each variety were greatly increased by the first and second increments of fertilizer at each farm each year (mean increases of 5·5 and 5·2 t/ha respectively), but less by the third (2·1 t/ha); total yields were increased from 35·7 t/ha with 1260 kg/ha of fertilizer to 48·5 t/ha with 3140 kg/ha. As yields increased the proportion of small tubers (< 44 mm) decreased, whilst that of medium (44–70 mm) and large (> 70 mm) tubers increased.

The removals of N, P and K by each variety of potato were calculated; for those given 2510 kg/ha of 13·13–20 fertilizer, the mean amounts (kg/ha) removed were 154, 20·6 and 242 of N, P and K respectively, or 47, 14 and 58% of the amounts given. A balance sheet showed the amounts remaining in the soil after potato harvest.

Winter wheat was grown after the potatoes to value residual N. In 1972, when average amounts of winter rain fell, grain yields were increased by these residues on the heavy Rothamsted soil, but not on the sandy Woburn soil. In 1973, after an unusually dry winter, the residues consistently decreased yields because of lodging, showing that little N had leached during the winter. % N in the grain was increased by the fertilizer residues each year.