Experiments to study the effect of date of sowing at two centres in each of 3 years, 1967–9, are described and the results discussed. The earliest sowing date (early March) gave the highest yield of spring barley at both centres and of spring wheat at Sutton Bonington, but at Boxworth in 1967 and 1969 later sowing (early or late April) gave higher yields of grain, which were attributable mainly to increased grain numbers per ear. It is argued that in 1968 poor weather conditions in July and August resulted in very low 1000 grain weights, and although grain numbers were again higher from late sowing at Boxworth this treatment gave the lowest yields because of poorly filled grain. The date of sowing also affected yields of total dry matter, ratios of grain to straw, leaf-area indices, numbers of grains per ear and 1000 grain weights.